Laser Interferometer Gravitational-Wave Observatory

LIGO Caltech 40 meter prototype

WEEKLY REPORTS

September 10, 2009

• In recent weeks we had great visits from Jan Harms (U Minnesota) and Yoichi Aso (U Tokyo). Jan showed us how to set up a quick seismic station to do seismic huddle tests for our Adaptive filtering seismic sensors. We now have 2 Guralp CMG40-T, 7 Wilcoxon 731A low frequency accelerometers, and one Nanometrics Ranger SS-1. The huddle tests allow us to measure the sensing noise limits of the sensors.
• We also have a prototype 3-frequency EOM which was put together by SURF student Stephanie Erickson. It uses a commercial New Focus KTP phase modulator and an external RLC network. The complete package makes 3 resonances which have a 50 Ohm impedance on resonance. Tests with an OSA verified the phase modulation response of the EOM.
• We had a failure of the APC 2200W UPS that powers the 40m vacuum rack. It failed suddenly after 15 years. We purchased the modern version of the same model from APC and have everything running again. No vacuum hardware was damaged by the UPS failure and our interlocks worked as expected.
• Peter King came over and upgraded our PSL Reference cavity temperature box to the LHO version (documented by H Radkins). This is to enable measurements of the AC temperature fluctuations of the vacuum can so as to estimate the frequency fluctuations in the 0.1-10 Hz band for AdvLIGO lock acquisition estimates. During this process we became disgusted with the ancient MINCO PID controller that is used in our PSLs and we have replaced it with a BNC cable and a Perl script. This software based approach has a lower noise and is easily tuned up from the control room.

July 30, 2009

• Wiener filtering: moved accelerometers around to explore changes in wiener filter efficacy with the MC.
• Relay out of ETMY oplev path to reduce acoustic sensitivity.
• More work on tuning up a tri-mod EOM circuit

July 17, 2009

• Work continues on design of the tri-mod EOM, the acoustic noise cancellation, and the GigE phase camera.

June 25, 2009

• Work continues on measurements of the DARM loop and calibration of it. This is going to require automation of determination of the readout quadrature.

June 18, 2009

• We have four SURF students starting work at the 40m this week.
• Work continues with characterization of the DR interferometer: Calibrations and diagonalization of the sensing and actuation matrices.
• Testing ongoing of the 35 W eLIGO beam dumps: BRDF and assembly.

June 4, 2009

• After the MOPA chiller replumbing, all of the standard laser trends are very stable.
• Work has gone on with tuning the handoff between single demod and double demod locking.
• Held a meeting of the 40m TAC on Thursday. Slides are in the wiki.

May 28, 2009

• This week has seen MOPA laser troubles.  After last week's efforts with the frozen needle valve in the NPRO cooling line, we noticed condensation perilously near the NPRO circuit board.  The water temperature was at 6C.  We decided to leave the MOPA off over the holiday weekend, and upon return, the chiller was no longer chilling.  We are searching for an available, working, spare chiller. An AC repairman thinks that we may be able to repair a flaky capacitor inside the chiller and get the compressor started again. 
• We also noticed that it's confusing how to hook up the water chiller lines to the MOPA, because of the labeling.  Reversing the flow means the needle valve in the NPRO line is unclosable (it only works in one direction).
• Work is continuing on wiring up the new molecular turbopump for a smooth transition.
• Calibrating the LM34 temperature sensors has proved problematic.  According to a set of commercial sensors, the LM34-based sensors were not all at the same temperature inside the "cool house."  Moreover, the PEM ADCU (ICS110-B) channels are not calibrated, differing from the nominal 16384 cnts/V by as much as 2.4% in the one measured case (as measured against a pair of Flukes, spec'd accurate to ~0.09%).  So the LM34 sensors will be calibrated one at a time, using the commercial sensor as a common temperature reference, and a Fluke as a common voltage reference.

May 21, 2009

• The ETMY coil balance looked bad. This was noticed by leaving the arms locked for a long time and monitoring the transmitted power change. The Y arm power slowly drifted and it had a very good correlation with the DC feedback force to the ETMY coils. If the coils are not balanced, the DC force meant to be applied to the beam direction also causes rotation of the mirror. By running f2pRatio script, which calculate the length to pitch coupling using the oplevs, we re-balanced the ETMY coils. Now the length to pitch coupling is much smaller for ETMY.
• When running the f2pRatio script, we found that the ETMY oplev was too noisy (about a factor of 10 worse than the others).  There was a lens mounted on a skinny long rod in front of the QPD. This lens induced a lot of seismic noise coupling to the ETMY oplev. We removed the lens.
• Caryn has been measuring the relative drift of the temperature sensors by putting them in a cooler box and recording the temperature readings with spare PEM channels. She observed some data drops in the recorded data. We don't know why at this moment.
• With the fclose problem fix suggested by Chris Wipf implemented, the seisBLRMS script has been working for 3 weeks without a crash.

May 14, 2009

• The DARM response of the DRFPMI detuned to non-springy side was measured with DC readout. It shows an RSE peak at 3.8kHz and the optical spring stiffness below 70Hz as expected. The measured response has an extra small bump around 2.5kHz and a steeper cut off above the RSE peak. The reason for this is still under investigation.
• The arm losses were measured by taking the ratio of the reflected power between the locked and unlocked states. The measured losses were:
  XARM: RTL= 171 (+/-2) ppm
  YARM: RTL = 181 (+/-2) ppm
  These values indicate the arm finesses to be about 1200 if we assume the ITM transmissivity to be the designed value. From the arm power transmission curves measured for the arm absolute length measurement, the arm finesse was also estimated to be around 1300. So these are roughly consistent.
• The laser power started to fluctuate a lot in a daily cycle after the cooling water pipe was unclogged. The water chiller temperature is a few degrees higher than it used to be. It looks like the chiller is overloaded.
• The MC started to be kicked out of lock intermittently. This is probably caused by the MC3 side coil driver being intermittent. This problem is intermittent, and it has been quiet for a day now. We will see if it happens again.
• We replaced the ETMY optical lever quad as it was noisier than the others.
• We installed a new 14TB RAID for the framebuilder to have longer data lookback.

May 7, 2009

• This week the locking of the 40m became fairly regular. The initial acquisition is still done with the a common mode offset in the arms. During the reduction of the CARM offset, the error signals are handed off and the DARM demodulation phase is continually adjusted to account for the rotating carrier field from the CARM detune.
• Currently the common mode servo's MCL handoff succeeds ~2/3 of the time and the switching of the CARM error signal from REFL_DC to REFL166_I succeeds ~1/2 the time, so there's a total success rate of 30% for lock acquisition.
• The lock is also currently happening on the anti-spring side and the dark port image is not very clean (left hand monitor in this elog entry:  http://nodus.ligo.caltech.edu:8080/40m/1561). Analysis of an OMC mode scan is ongoing, but it looks messy.
• Next up is to stabilize the acquisition some, do some noise coupling measurements, and then try for the SRC offset that gives the standard optical spring configuration.

April 30, 2009

• The interferometer finally locked fully with the RF signals. The trick was to continuously adjust the DARM (AS166) demodulation phase as the CARM offset is reduced. This is because the CARM offset rotates the carrier phase at the AS port whereas the local oscillator (166MHz SB) is not affected. With the tuning of the DARM demodulation phase, we are now able to maximize the arm power (i.e. the arms are fully resonant) robustly. Currently the CARM offset does not go to zero when the arm power is maximized. This is because there is some offset in the CARM error signal. The origin of the offset is probably a wrong demodulation phase. Rob's Optickle simulation suggested that it is hard to find the optimal demodulation phase for POX_1I because it is different for X-arm single lock and for CARM. Therefore, we decided to use REFL_2I as the final CARM signal, which was actually the original plan. The REFL_2I demodulation phase can be optimized using SRCL with DRMI configuration. We are now working on this.
• The spare turbo pump turned out to have an O-ring flange, which we don't want. Osaka vacuum kindly offered to exchange the pump with a new copper-gasket type pump. It will arrive next week. Until then, we are pumping with the cryopump. Joe is preparing for the switch over to the new pump controller.

April 23, 2009

• Lock acquisition effort continues. The AutoDTT script from LHO was modified to suit the 40m environment. It indicates that DARM becomes unstable at around arm power = 30. The phase bubble of the DARM shrinks at this power level. We are adjusting the filters to compensate for this.
• Jenne replaced the mixer for the PMC demodulator because it was level 23 and our current LO level is 16dBm. This is the same problem seen at LLO and LHO. The guess is that the ERA-5 in the PMC Frequency Reference card is dead. The new mixer is level 13 which is easily sufficient and the PMC locks fine with the new mixer.
• The main turbo pump failed on Wednesday. It did an emergency landing on the dry bearing. With this incident, the number of the emergency landing exceeded the allowed limit (5 times maximum). We have to replace the bearing but it is hard to find a replacement because the pump is more than 15 years old. We will have to replace the old turbo with the spare one. Since the spare pump is newer model, the EPICS interface and the interlock wiring must be updated. We are preparing for this. Meanwhile, the vacuum is maintained by the cryopump.
• Joe and Peter made a skeleton model of a suspension controller with the advanced LIGO style realtime code generator. The idea is to make highly advanced screen / FE so that this can be used as a template from which to copy from for the triples and quads. This will also be the control for the 40m upgrade SOSs and MOSs.

April 17, 2009

• Using REFL_DC as CM error signal, we don't see the 3.8kHz problem and the loop gain can be higher than with the PO_DC. We are still adjusting the CM gains as the power is ramped up. However, the IFO loses lock randomly. REFL_DC may be more sensitive to the alignment of the PRM and other mirrors. We are considering to stiffen the optical levers. To investigate the cause of the lock loss, we have imported AutoDTT scripts from Hanford. At this moment, we can go up to the arm power of 26.
• PMC local oscillator level was found to be about 16 dBm where the mixer of the PMC demodulation board is level 23. We ordered a level 13 mixer to replace the old one as was done at LLO and as LHO should be doing right now. Jenne edited the PSL EPICS database and screen to show the LO level in "dBm".
• We tested a new wireless headphone (Amphony 2500), which is digital as opposed to our old analog headphone. It turned out to have a very good sound quality and runs at 5.8 GHz so its doesn't get 802.11 traffic. The noise is almost unnoticeable within ~20m range from the transmitter. This is very useful for acoustic noise hunting and should be bought for both sites.
• Alberto generated several plots of PRCL error signals as a function of CARM offset for REFL1I, REFL3f, and REFLDD. He is now compiling the results into a suggestion for the current and upgraded 40m locking.

April 10, 2009

• Optickle simulations suggest that the disturbing 3.8kHz peak in the CARM loop comes from the fact that PO_DC is also sensitive to the DARM. The 3.8kHz peak is the DARM RSE peak. The simulation suggests that REFL_DC is less vulnerable to the DARM coupling. Therefore, we are trying to use REFL_DC as the CARM error signal while ramping up the power. So far, we successfully handed of the CARM error signal to the REFL_DC. We confirmed that there is no 3.8kHz peak in the AO path. Now we are optimizing the loop gains during the power up.
• Rana demonstrated to lock X-arm by feeding back to the MC2 (no ETMX/ITMX feedback). It was stable and the lock was fairly quick.
• The Guralp seismometer is finally back from repair. We will use it for the adaptive noise cancellation.
• Rob recompiled the LSC code with the fix for the denomalization problem provided by Alex. It reduced the load of the LSC computer, which had been overloaded occasionally.

April 3, 2009

• The MC1 mirror was still drifting a lot. The alignment sometimes got sudden jumps and the MC lost lock frequently. We put monitor probes on the MC1 coil driver and looked for the cause of the glitches. During the process,Yoichi found that a cable connecting the coil driver to the satellite amplifier had a broken connector on the satellite box side. He replaced it. Since then the MC1 has been quieter than before.
• Lock acquisition work: The hunt for the cause of the 3.8kHz peak in the CARM loop continues. We locked the CARM by only feeding back to the ETMs and increased the arm power to 4. Then measured the transfer function from CARM (excitation to the ETMs) to PO_DC signal. This measurement involves no feedback/actuation to the laser frequency but we still found the 3.8kHz peak. This suggests that the peak is a real optical response of the interferometer. 3.8kHz sounds like the DARM RSE peak. Indeed, the peak is present in the DARM loop to. We are trying to identify how the DARM RSE peak is coupled to the CARM loop by adjusting offsets in the various places.

Mar 27, 2009

• We tried to lock the mode cleaner without the frequency stabilization servo. To do so, we run a cable from the servo output of the MC board to the test input of the FSS box to drive the laser frequency directly. With an addition of a passive lead-lag filter to compensate for the difference of the cavity poles between the reference cavity and the MC, we were able to lock the MC with 80kHz bandwidth without the FSS. We are trying to optimize the feedback filter to increase the bandwidth further. Single arm lock is possible with the new MC locking scheme. But the frequency noise at the low frequency is larger and the arm power fluctuates a lot. Adaptive noise canceling should be able to relief this problem.
• The locking is still obstructed by the 3.8kHz peak in the CARM loop. A new PD was installed at POB. We used to use the DC output of the SPOB PD but it has a very small aperture and low signal level. 3.8kHz peak is still there with the new PD. This exclude the PO_DC PD from the cause of the problem.
• Kakeru calibrated optical levers with OSEMs to check the consistency with his previous calibrations using the cavity power change. The results were consistently smaller than the cavity power method by a factor of 3 to 4. Seems like there are systematic errors.

Mar 20, 2009

• The drift of the MC1 suspension that we have noticed for a long time was finally tracked down to a loose cross-connect connection between the 4116 DAC and the MC1 coil driver.
• Kakeru did calibrations of optical levers by looking at the single arm cavity power v. optical lever signal and fitting. This operation has been automated so that it can be ported to the sites and used to bootstrap the WFS sensing matrix calibration. He calibrated all the core optics and the calibration factor are in the order of 100 urad/cnt. The error of the calibration factors is estimated to be 30%.
• For lock acquisition, we are still fighting against the peak structures around 3.8kHz in the CARM loop. Transfer function measurements of the Additive Offset path in various configurations including single X-arm lock imply that the peaks are not present when the AO path loop is not closed. As the AO path gain is increased the loop shape changes and the peak grows. With fine tuning of the DARM and CARM gains during the power up, we can robustly reach arm power = 20 or more, but the 3.8kHz oscillation eventually comes in and breaks lock.
• seisBLRMS (a MATLAB based data monitoring tool to compute band limited RMS of seismic channels and put the resultant values into EPICS channels) was compiled against version 2008a, which used to fail. It is now running continuously. We now have good trend channels for seismic activities and should be able to get long term data for the construction and water flume work.

Mar 13, 2009

• The lock acquisition process now goes up to arm_power = 27 (where the single arm power = 1). Around this power level, two problems are observed. One is an oscillation of both DARM and CARM loops at 3.8kHz. Actually there are many broad peaks in the DARM and CARM error signal around kHz region. Those peaks are also visible in the single arm lock error signal (both arms). The 3.8kHz peak grows when we engage the fast CARM feedback, and sometimes kicks out the IFO. We twiddled the gains of the FSS, PMC, MZ, ISS and MC loops but no change in the peaks was observed. Although the shape of the peaks looks like acoustic origin, tapping on the tables, chambers also did not change the peaks. A historical spectrum of the X-arm error signal provided by Osamu shows the same peaks in 2003. The behavior of the noise is consistent with this being MC mirror displacement noise, RFAM on the MC sideband, or jitter in the Piezo Jena steering mirrors after the MC.
• Kiwamu and Kakeru installed a new QPD for IOO-ANG to monitor the drift of the beam pointing into the MC.
• Matt compiled a new version of tds tools for 40m. The strange data glitch problem of the TDS tools is fixed; we were running a 2 year old version of the TDS-DAQ code.
• The CLIO digital system test setup was found to be interfering with the 40m data acquisition system, after some debugging. It was moved out of the 40m lab area to the office area to avoid further interference with the 40m system.

Mar 6, 2009

• The lock acquisition process still has difficulty in passing the arm power of 10. In this state, all the short degrees of freedom are controlled by the DD signals, and the DARM is on the AS RF signal. The CARM loop is controlled by the PO_DC with the high bandwidth AO path engaged. When ramping up the arm power, the IFO loses lock suddenly around the power = 10. Every time it loses lock, the power in both of the arms goes up rapidly up to 30. It always goes up not down. This seems to be the key to the problem, but the cause of this is still not understood.
• Joe has been adding useful functionalities to the GigE camera softwares. He is preparing for a clean release of his SnapPy package for people to use. He also received two more cameras and set up a dedicated network between the cameras and the server computer. It was confirmed that one gigabit network card with a gigabit switch can handle the four cameras at the same time with the maximum image size.
• Osamu and Kiwamu installed two SuperMicro computers to test the advanced LIGO style digital control system. They are now installing various softwares, such as real time kernel, the real time code generator and so on.
• Jenne is trying to do the winner filtering noise subtraction for the S5 data.
• Alberto is working on the simulation of the double-demod signals for the current 40m to compare them with the 3f signals.

Feb 27, 2009

• We are optimizing the lock acquisition process. Now we can robustly reach to the point where the CARM error signal is handed off to pirck-off DC signal. The IFO loses lock after the arm power is ramped up to 8 (normalized with the single arm lock power). We are continuing to optimize the servo while the arm power is increased.
• During the lock acquisition work, we found a poorly soldered resistor in the MC board, which caused a problem in the AO path. We fixed it by re-soldering this resistor.
• JoeB made nice python bindings/wrappers for the GigE camera code. He also wrote simple server and client programs using his python bindings to get images from a camera and broadcast the video in a compressed format. His approach uses gstreamer framework, which makes it extremely flexible. Currently we have a camera set up to watch one of the video monitors in the 40m control room. With simple commands, we can watch this image with reasonably high frame rate even at home. So we can now fully operate the IFO from home ! His program can do centroid detection, and feedback to EPICS channels. It also has an automatic exposure control. He is almost ready to ship the technology to the sites and replace Spiricon.
• Caryn calibrated her temperature sensor. She found that the MC chamber temperature is consistently higher than the air temperature around. She made a new power supply for the sensor. It is less noisy.
• Kakeru found that tdsdata returns data with glitch samples. This problem is reproducible on Solaris, Linux 32bit and Linux 64 bit.
• Alberto is working on the simulation of 3f locking and DD locking.

Feb 20, 2009

• After the MOPA power was recovered, the ISS became unstable at the lower UGF.
• We discovered that since the MOPA is operating with only 2.7 W output (after years of degradation), the mode matching within the rods is probably not good. Since there is no apparent problem with the electronics, we conclude that the PA current is modulating not only the 808 nm pump power, but also the mode matching within the amplifier.
• So the low frequency (~0.1 - 10 Hz) change in the current shunt transfer function is mainly due to the degraded pump diodes in this laser (the original H2 MOPA).
• A scan of the pumping current spanning 2A around the nominal pump current showed that we were in fact in a region where the output power decreases as the current goes up. Moving the nominal operation current of PA down by about 0.5A made the ISS a lot more stable.
• We will remove this bad idea from the 40m during this summer's upgrade and will replace it with an AOM as we have in Enhanced LIGO.
• After the ISS was fixed, we resumed night shifts for lock acquisition of the full DRFPMI.
• As the seismic noise from the construction of the nearby IT building has decreased, the installation of the large water flume on the other side of the wall from the 40m (in the old CES highbay) has begun. This is a new geology lab that will run mostly in the daytime - so far the contractor has been willing to cooperate with us so that the construction noise will start at 4 AM after interferometer work stops and will stop ~noon when the interferometer people straggle back into the lab.

Feb 13, 2009

• After the MOPA power was recovered, the ISS became unstable at the lower UGF.
• We discovered that since the MOPA is operating with only 2.7 W output (after years of degradation), the mode matching within the rods is probably not good. Since there is no apparent problem with the electronics, we conclude that the PA current is modulating not only the 808 nm pump power, but also the mode matching within the amplifier.
• So the low frequency (~0.1 - 10 Hz) change in the current shunt transfer function is mainly due to the degraded pump diodes in this laser (the original H2 MOPA).
• A scan of the pumping current spanning 2A around the nominal pump current showed that we were in fact in a region where the output power decreases as the current goes up. Moving the nominal operation current of PA down by about 0.5A made the ISS a lot more stable.
• We will remove this bad idea from the 40m during this summer's upgrade and will replace it with an AOM as we have in Enhanced LIGO.
• After the ISS was fixed, we resumed night shifts for lock acquisition of the full DRFPMI.
• As the seismic noise from the construction of the nearby IT building has decreased, the installation of the large water flume on the other side of the wall from the 40m (in the old CES highbay) has begun. This is a new geology lab that will run mostly in the daytime - so far the contractor has been willing to cooperate with us so that the construction noise will start at 4 AM after interferometer work stops and will stop ~noon when the interferometer people straggle back into the lab.

Jan 29, 2009

Upgrade Work

• We are continuing to buy equipment for the upgrade and analyzing parameter values in the aux. optics.
• We have decided to deviate from the nominal AdvLIGO modulation frequencies of 9 & 45 MHz and to use 11 & 55 MHz instead - the larger frequencies will allow for shorter lengths greatly simplifying the optical layout.

Other

• Alberto took the first measurement of the PRC absolute length by injecting his laser from the dark port. He found a peak at the expected frequency but the shape is somewhat different from what was expected. He will take more measurements.
• The MOPA output power started to decrease about a week ago. We found that it was caused by a clogged pipe for the chiller water going to the base of the NPRO. We tweaked the needle valve, which seemed to be clogged, and the water started to flow again. Now the laser power is back to what was used to be.
• After the laser power was recovered, the ISS servo started to be unstable. It is oscillating around 3Hz. Kakeru and Yoichi measured the open loop transfer function to find that the phase margin at the low frequency UGF was too small. Kakeru and Peter is now modifying the servo filter to give more phase margin.
• Joe tried to get data from our new RGA through an ethernet-RC-232C converter. Unfortunately the baud rate of the RGA is fixed and this particular baud rate is not supported by our ethernet-RC-232C converter. We have to buy a new converter which support the necessary baud rate.
• Megatron (our AdvLIGO style front-end computer) started to make huge fan noise a few days ago. Turned out to be similar to a problem with the Sun 4600's that is seen at Hanford - Joe and Dave Barker talked it over and solved this by power cycling the machine.

Jan 22, 2009

• The single arm lock instability was found to be caused by the mis-centered beams on the PO PDs. After centering those, the arm lock is now stable and the dither-alignment works well. We also found that the relative gain of the MC_L compared to the MC_F was a factor of two lower than it should have been. The MC_F gain was increased when we fixed the double path AOM, which had been single path for a while, a few months ago. Since then the cross-over frequency of the MC_L and MC_F was too low. We fixed it.
• Kakeru and Peter are working on the calibration of the optical levers using the arm cavity g-factors and arm power as a reference. They calibrated Y-arm oplevs this week. They will extend the method to the recycling mirrors and the BS.
• Alberto has done a simulation of the absolute length measurement of the PRC by injecting an auxiliary laser beam from the AS port. The signal seems to be small but he is trying this method to see if he can measure something.
• In the meeting, it was decided that we will use 11MHz and 55MHz modulations in the upgraded 40m instead of 9&45MHz. The reason behind the decision is that the 9MHz SB requires too long power recycling cavity. Yoichi struggled for some time with the 40m optical layout CAD model and found that it is next to impossible to accommodate an 8m long PRC in our chambers using 2 or less folding mirrors. A bonus coming with the 11MHz base frequency is that we don't have to change the MC length.  Yoichi is working on more detailed optical layout with 11MHz SB including PO beams and optical levers.

Jan 15, 2009

• We decided to resume the 3f locking effort in parallel with the current lock acquisition scheme. For the upgraded 40m, 3f locking is probably necessary. For the current IFO configuration, the benefits of 3f compared to DD is not clear at this moment. Alberto is running simulations to check this.
• Kakeru made an auto-updated webpage showing recent screen shots of important MEDM screens.
• We re-optimized the demodulation phases of the DD signals for the hand-off from single demod. signals. This was necessary because we restarted the DD oscillators after the power failure.
• Recently (past few days), the MC has been kicked out of lock occasionally (twice a day or so) without any apparent seismic activity. Also the single arm lock has also been unstable these days. These two problems may be correlated as the arm instability seems to be caused by increased frequency noise. We are trying to track down the source of the problems.

Jan 8, 2009

• The finesse of the frequency reference cavity of the PSL was measured by the cavity ring down method. The incident light to the reference cavity was shut down by a high-voltage pulse driven pockel's cell and a PBS. Although we used high-voltage compatible cables and put a slow-down resistor in series to the pockels cell, the light power exhibited ringing after the shut down. So a simple exponential fit was impossible.
• The cavity pole was estimated by fitting the measured cavity transmission power with the simulated response of a first order low-pass filter to the incident power change. The result of the cavity pole measurement is fc=47.6kHz(+/-0.6kHz). The design was 35kHz. It was 38.74kHz last time measured (years ago). This means that the round trip loss in the cavity has increased from 600ppm (design) to 800ppm (current).
• The mode cleaner alignment has been drifting a lot for unknown reason for a few days. We re-centered the MC WFS QPDs. We also re-activated the QPD to monitor the PSL beam pointing to see if there is any drift in the PSL beam.
• We had a power failure on Saturday (20th) morning. This was due to the Caltech IT people making a mistake while doing network maintenance on a nearby building.
• The main valve for the turbo pump was automatically closed and the laser was turned off. All the FE computers lost the power. After the power recovery, the vacuum pump and the laser restarted normally. The power supply of an RFM bypass box was found to be broken. We replaced it with a lab. power supply for the moment. After fixing several computer problems, the 40m interferometer got back to normal by Wednesday (Dec. 24).
• The optimal modulation frequency for the MC transmission was measured for the first time after the MARCONIs were locked to the Rubidium clock, which is itself locked to GPS via a 1PPS signal. After this, the AS166 demodulation phase was adjusted by maximizing the MI signal.
• A script to take screen shots of MEDM screens in background (i.e. without a physical display) was written and is running every 5 minutes as a cron job on nodus. It posts the screen shots to a web site so that people can check the status of the lab quickly from outside.
• Kakeru, a visitor from Japan, is now making a web page to show those images nicely. Alan also wrote a script to create a table of EPICS channel values and post it on a web page. This script is also running as a cron job.
• A perl version of ezcademod was written because the compiled version does not run on linux currently. Dither alignment scripts were modified to use the perl version and now they run on linux machines.
• Had a meeting of the 40m TAC. Review report is being compiled by Ken Strain.

Dec 12, 2008

• After the MZ and MC alignment work, the drum head modes of the MC mirrors became visible in the noise without artificial excitation. The MC error signal is down converted by a lock-in amplifier and sampled into the DAQ system. Caryn is working on making a monitor program to track the frequency change.
• Alberto and Caryn are working on the calibration of the RF-AM monitor after the MC. For the first step, they calibrated the "Janne laser" (a small diode laser with RF amplitude modulation capability). They found a strange notch in the transfer function. Also the transfer function had ripples due to reflection from impedance mismatch. They are now using the beat from the PLL system for the abs. length measurement for the calibration.
• A mag 5.1 earthquake on the Friday night (Dec. 5) tripped most of the optics. Inspections by free swinging spectra observed no major damage. However, the earthquake seemed to kick the ITMY-UL OSEM back to normal, where it had been outputting zero for a few weeks. Now the IFO mirrors are re-aligned and locks in DRMI.
• The rough pump for TP2, which is the foreline pump for the main turbo, failed on Monday. It was replaced with a spare.

Dec 3, 2008

• This week the MC was re-aligned after doing the MZ work and the RFAM minimization. This somehow reduced the noise around 28 kHz, allowing the MC drumhead modes to become visible in the MC_F channel with only a 10 second integration. This will make the drumhead mode tracking and absorption measurement easier.
• The MC Wiener filter based noise subtraction was enhanced by introduction of another seismometer. Now using 6 Wilcoxon accelerometers, one Ranger SS-1 seismometer, and a Guralp CMG-40T, we are able to get more than a factor of 10 reduction in the MC_L feedback channel from 1-25 Hz. Noise enhancement out of band is fairly minimal. We are starting to write a noise subtraction paper. Next is to resurrect the adaptive, online noise subtraction system to see how well we do over very long time scales with this setup.
• The coherence among separated seismometers agrees well with the well known Bessel 2-point correlation function.

Nov 13, 2008

• The Ranger seismometer had a bad connection to the DAQ. This was repaired.
• The performance of MISO Wiener subtraction for removing seismic noise from the Mode Cleaner was evaluated. It seems that this feedforward system would not need to be adaptive. There is no observed degradation in the residual noise spectrum over ~5 hours with a static set of FIR filters.
• The prototype LM34 based BSC temperature sensor was installed on the MC1/MC3 table. We will now start monitoring the trends and also the drumhead frequencies of the MC optics to make a long term absorption sensor. If this seems OK, we will fabricate more of these to install at the sites for the science run; one for each test mass chamber.

Oct 31, 2008

• Alberto found that the reason why he could not measure TEM01 mode with his absolute length measurement setup -- there was too much clipping of the beam by a knife edge. After adjusting it, he was able to measure the split of TEM01/10 mode resonances. The measured ROCs are, R_horizontal = 56.26 0.01 m, R_vertical= 56.63 0.01 m, where the design ROC is 57.57m. The astigmatism is about 0.37m.
• Calculation of the location of higher order mode sideband resonances with respect to the main resonance was done based on the measured g-factors. There is no accidental coincidence of HOMs with the main resonance found.
• Caryn and Alberto made and installed a temperature sensor on the MC chamber for the MC1 and MC3 mirrors.
• Jenne measured the noise of the Guralp seismometer box. Below 3Hz, the noise goes up quickly. The velocity equivalent noise at 1Hz is 1e-9 (m/s)/sqrt(Hz). She will compare the noise with the normal ground motion level of the 40m lab to see if the noise level is acceptable.
• Peter calibrated the FSS LO level monitor and now the EPICS channel shows the LO level in dBm.
• Steve installed lexan plates for all the view ports facing upwards to protect them from accidentally dropping something on it.

Oct 24, 2008

• An AD797 (U10 in the PC path)on the FSS board was found to be oscillating. It was replaced with an AD829. Once a phase compensation capacitor was implemented, the oscillation was mitigated. We also replaced another AD797 on the board with an AD829 just in case. During the work, to see the stability of the op-amps, a LISO model of the FSS circuit was made. It can be useful for noise estimation, stability analysis, etc. It can be found in the 40m SVN.
• Currently, the FSS UGF is about 400kHz (before the recent FSS works, it was less than 200kHz). With the increase of the FSS UGF, now the MC open loop transfer function does not have a bump at around 100kHz, which used to limit the bandwidth of the MC loop. Currently the MC loop's UGF is about 70kHz. This should now re-enable good CM servo action for interferometer locking.
• Steve measured the divergence of the new HeNe laser for ITM optical levers. Then Yoichi calculated necessary lens focal lengths and locations for focusing the beam on the QPD.
• Alberto wrote a python script to run the absolute length measurement automatically. This entails controlling the HP RF analyzer (8591E) and the IFR (previously Marconi) 2023A that is in use at the sites; by purchasing a few of these cheap GPIB-Ethernet boxes we should be able to make some of the laser noise couplings for Enhanced LIGO automated.

Oct 17, 2008

• After tuning the gains and phases of the control loops, the stability of the DRM lock to +f2 side band improved a bit. But it is not stable enough. Will try direct locking by the double-demod signals.
• The correlation between the EMI around the master oscillator of the MOPA and the FSS error signal was measured by picking up the EMI by a handmade loop antenna. There is strong correlation between the two signals at the harmonics of 78.1kHz and many other lines.
• Peter measured the optimal demodulation phase for the FSS and made a cable to realize the necessary phase delay.
• Alberto started to port the GPIB program for the absolute arm length measurement written by Koji in LabView to a python script with an interface from EPICS.
• A new control room computer running 64bit CentOS was installed. We figured ways to run control room tools (dataviewer, dtt, medm, etc) on 64bit linux.Our another 64bit linux machine, which could not run control room tools, is now capable of running those tools too.

Oct 10, 2008

• Lock acquisition: After the locking of DRM on the springy side (+f2 sideband) became almost stable, it went unstable again for unknown reason. -f2 side is still stable. Will tweak the sideband frequencies.
• Rich brought back a replacement DC-DC converter for the MOPA NPRO. We found that our NPRO is older than the ones at the sites, and it requires to move the laser to install the replacement DC-DC converter. We postponed to do it. Coherence measurements between the FSS error signal and the Sorensen power line (+24V) show (almost) no correlation between them.
• New HeNe lasers for ITM optical levers arrived. Their spec is 10mW and the actual output is 15mW. Steve is installing them.

Sep 26, 2008

• Rob optimized the output matrix for the short degrees of freedom (PRC, SRC, MICH) to make the lock of DRM+2ARMS quick and robust. The arm offset is reduced down to zero yet. The DRM is locked in the anti-spring configuration still. The MICH->PRC matrix coefficient is 3 times smaller than it should be. There might be something wrong with the PRM actuation or sensing.
• The PMC loop bandwidth is found to be now limited by a body reasonance at 18kHz. We will add a notch at this frequency too.
• Alberto is working on the TEM01/10 mode split measurement of the X-arm. Currently the arm lock to 10 mode is not stable enough to do the measurement.
• The new RF oscillator for the FSS was packed into a box and installed. It is working fine so far. Noise measurements of the oscillator are planned.
• The GPIB to ethernet converter arrived. It was configured to connect to the 40m network through an ethernet to wireless bridge. We can now communicate with a GPIB instrument connected to the converter using telnet. A python script to automatically download data from any network connected spectrum analyzer (SR785 or Agilent 4395A) was written. We will order more of those GPIB controllers.
• The FSS error signal was calibrated by sweeping the laser frequency and measuring the PDH signal slope. The result was consistent with the previous calibration using closed loop transfer functions.

Sep 19, 2008

• Lock acquisition work resumed. Rob was able to lock each component (Arms, PRMI, DRMI) but the 2Arms+DRMI lock has been unsuccessful so far. He was trying to lock on the spring side. Will try no-spring side, which is easier to lock.
• Alberto measured the X-arm length (38.4580+/-0.0003m). He is preparing for the TEM01/10 split measurement.
• Yoichi calibrated the FSS error signal. The in-loop frequency noise is about 10^-2 Hz/sqrt(Hz) from 100Hz to 10kHz. Seems to be fine but be aware that this is an "in-the-loop" measurement. A canonical cavity pole frequency of 38kHz was used for the calibration. A cavity pole measurement with cavity ring down is planned for more accurate calibration.
• Joe set up a DNS server for the martian network so that we don't have to maintain/etc/hosts file on each host.
• Wenzel oscillator and a mini-circuit RF amplifier arrived. Peter connected them and tested it with the FSS overnight. They worked fine. He is now putting them into a shielded box.
• A GPIB-ethernet converter and an ethernet-wireless bridge arrived. Yoichi is setting them up.

Sep 12, 2008

• The mode cleaner has been working fine for a few days. The POY RF amplifier, which was temporarily used for the FSS LO amplifier, was put back to the original duty, so that the lock acquisition experiment can be resumed. A Wenzel oscillator arrived from Hanford. Peter will put together a new 21.5MHz reference for the FSS once we receive an RF amp, a splitter, attenuators etc. from mini-circuits.
• We found loose cables for -24V supply in the IOO rack. The ISS had been behaving strangely (saturates easily). After fixing the power supply cable, the ISS seems to be working fine.
• The absolute length measurement was resumed by Alberto. The PLL is now working. The NPRO beam and the dark port IFO beam were aligned. Beating between the injected NPRO and the IFO beam was observed at the Y-end when the NPRO beam resonates in the arm.
• We are designing automation of the measurement process by developing network connected GPIB/EPICS interfaces.

Sep 5, 2008

• The RF oscillator signal for the FSS was too noisy and the LO power was too low. The waveforms were distorted and there was much feedthrough of switching supply noise. We temporarily swapped it with a combination of a Stanford signal generator and a mini-circuit RF amplifier. Peter Kalmus is designing a new RF oscillator box.
• Now the UGF of the FSS is about 250kHz, a factor of 2 improvement. By setting the relative gain of the PZT path a bit lower than before, the MC finally locks for long stretches of time (easily more than 10 hours). However, we are still unable to increase the FSS UGF up to desired 500kHz. Accordingly, the MC UGF is still low (40kHz). We see many huge 77kHz harmonics in the error signal of the FSS. We think these peaks are limiting the slew rate of one of the op-amps in the FSS servo box. We plan to move the suspicious DC-DC converter near the NPRO crystal to outside of the NPRO box.
• PMC open loop transfer function was measured by Jenne. Now the UGF is 630Hz with the phase margin of 53deg. The bandwidth is still limited by some resonance. Lead brick configurations continue.
• Joe and Eric used the GigE camera to monitor the position of the beam spot on the ETMX and succeeded to move it by a servo to the ITMX. According to Joe's calculation, the camera can measure the position of a beam with a half micron precision if the beam is directly hitting the camera. By monitoring the beam position change by mis-alignment of the mirrors, we should be able to detect a fractional power loss of 2*10^-10 in the 40m arm cavity.

Aug 29, 2008

• After the pumpdown, work has mainly focused on understanding the instabilities in the MC.
• We have upgraded the PMC electronics to the version used at the sites (better IF filtering, more DC gain, etc.) and have installed a passive damper on top of the PMC body to reduce the height of the body mode peak.
• The instability seems to be coming from the FSS - we are now working to characterize the loop shape and to look to see if there are frequency glitches coming from the NPRO.

Aug 22, 2008

• After the recovery works from the earthquake in the air, we finally pumped down the system again on Friday. The vacuum is now fine, and the suspensions are all in ok state. Rob recovered the alignment of MC, both arms and DRM. So all the mirrors are aligned ok, although final tune is still necessary. During the recovery process, Rob developed a lazy but handy method of finding a good alignment of a cavity. That is, slowly modulating the alignment offsets of the relevant mirrors with triangular waves of different frequencies. Leave it for an hour or so and then look for times of cavity flashing in the trend. By reading the offsets applied at those times, the cavity alignment can be brought close enough to the optimal for final hand tuning.
• The mode cleaner looses lock sporadically, once or twice an hour.No apparent correlation with seismic noise. We are investigating the cause of this phenomenon.
• Diagonalization of the OSEMs of the repaired PRM was done by Jenne. An automatic diagonalization code is under development.
• For the adaptive noise cancellation, the accelerometers were calibrated so that they generate the same signal in response to the same acceleration input.
• Eric modified ezcaservo to be able to set the max, min, slew rate limits. He also wrote Matlab functions to convert the pixel position in the GigE camera image to the real position.
• Large groups of Japanese high school students visited 40m.

Aug 15, 2008

• Efforts to recover the interferometer from the earthquake damage are continuing. After some unsuccessful attempts to level the crooked PRM inside the chamber, we pulled the suspension out and inspected closely. Then we found that one of the stand-offs was missing. What seemed like a stand-off was actually a guide rod. Gary sent us spare stand-offs from Livingston. A new stand-off was placed under the guide rod, and we balanced the mirror. The PRM suspension was put back to the BS chamber on Thursday. We restored the alignment of the mirrors using the optical levers. We plan to pump down on Friday.
• Joe is slowly switching the 40m network to a Giga bit one. The Gaussian fitting code of the GigE cameras was tested using the PSL beam. The center position and the waist size were measured with relative errors of about 0.05%.
• John checked the relative positions of the higher order mode sideband resonances to the TEM00 carrier resonance in the Y-arm cavity using the g-factors measured by Koji. This time, the split of the horizontal and vertical modes by astigmatism was taken into account. For Y-arm, there is no +166MHz sideband (used for the springy lock of the SRC) resonance found around the main resonance up to TEM55. The closest higher order mode was found to be TEM50 of -33MHz at about 5kHz away from the TEM00 carrier resonance.

Aug 8, 2008

• After the earthquake, we've been working on recovering the interferometer. We vented the 40m vacuum system on Monday. The venting went fine. We inspected suspicious optics, ITMX, SRM and PRM. Apparently, the vent freed the ITMX, which seemed to be stuck somewhere before the vent. SRM's side magnet fell off and was stuck to the LR magnet. We picked it up and moved the side OSEM to the other side where another magnet was intact.
• We used the laser for the absolute length measurement (see below) as a reference and aligned the ITMX, ITMY, BS and SRM.
• The PRM magnets are all there and the mirror is not touching anywhere. But the mirror is tilted in pitch a lot. The wire seems to be in the grooves of the stand-off, although the farther side is difficult to see. We suspected the wire is shifted a bit on the bottom of the mirror. We lifted the mirror very gently using the bottom earthquake stops to relieve the tension of the wire and pushed the mirror on the face using another earthquake stop screw to make the mirror upright.Then the bottom earthquake stops were slowly lowered so that the mirror sits on the wire again. By repeating this many times, we hoped that the mirror comes back to upright position. So far we were able to change the angle, but the PRM is still not perfectly upright.
• The absolute length measurement of the Y-arm cavity was done by Koji. The FSR and the length of the Y-arm are measured to be: f_FSR = 3878678 Hz +/- 30 Hz; L_yarm = 38.6462 m +/- 0.0003 m
  He also found a split in the resonant frequencies of TEM01 and TEM10 modes. These techniques should allow use to measure the length of the recycling cavities as well as get the astigmatism of the COC.
• Joe has been trying to use Gstreamer for compressing the data from the GigE cameras to avoid the saturation of the network.
• The MZ for the fiber noise measurement is reconstructed with better mounts. Testing an AOM for the noise cancellation experiment.

Aug 1, 2008

• We suffered from the 5.4 EQ on Tuesday which was ~20 miles SE of Caltech.
• Currently, the PRM, ITMX, and others are severely mis-aligned and probably stuck somewhere. We are comparing the pendulum eigenfrequencies before/after the earthquake to check for touching. Currently the MC and the Y arm can be locked.
• We plan to vent on the next Monday.
• The MC length was measured by injecting a 2kHz signal into the MC error point and moving around the 166MHz sideband frequency to minimize the demodulated signal at the reflection port (166MHz). The 166MHz sidebands were 6kHz off the resonance. We corrected the other RF frequencies (33MHz, 133MHz, 199MHz) accordingly.
• The arm cavity absolute length measurement made some progress. Koji found the pump diode current of the auxiliary laser was too low. By putting it back to the normal value, the beat between the PSL laser and the auxiliary one can be seen at a reasonable crystal temperature (48.5 C whereas the MOPA's NPRO is at 45 C). The PLL now works fine and we are ready to search for the FSR of the arm cavity. This setup can also be used to find the transverse mode spacing (g-factor measurement).
• After finding that we were using a wrong order beam from the AOM for the reference cavity of the PSL, we picked the right one and re-aligned the RC. Now the incident power is about 10 times stronger than before. We also installed new lens mounts for better mode matching. However, even with the increased optical gain, the FSS loop bandwidth is still limited to below 200kHz where we expect 500kHz. The transfer function measurements show the loop should be stable even with a higher bandwidth. But when the gain is increased further, the phase corrector path suddenly starts oscillating incoherently. We are investigating this problem; it may be that the switching supply noise around 80 kHz is the culprit. If so, we will have to move the DC-DC converter out of the NPRO head and into a seperate box.
• We replaced the PMC which had been in the PSL for > 6 yrs with a spare in the lab. Range and gain of the new one is much larger leading us to believe that the old one had a bad PZT for many years.
• The fiber noise measurement setup was moved to the Bridge. A Mach Zehnder with a 50m fiber arm is now working. Marha measured preliminary noise spectra.
• The adaptive filter now runs a new C code, which saves the FIR coefficients and restores them when invoked next time. A bit of adaptation time improvement was achieved.
• Max added 40m specific seismic noise estimate to the noise budget code.
• We got another GC750 CMOS camera. Joe found it has an identical "background" pattern in 1064nm as the one we already have. The camera to EPICS communication is now working and so the camera based cavity alignment can now begin.
• Eric did PMC scans to test fitting the mode peaks so as to make an automated HOM analyzer, similar to the one used on the diagnostic breadboard from LZH/AEI.

July 25, 2008

• The fiber noise experiment has been moved out of the 40M and into the sub-basement of Bridge to take advantage of the digital control system. Seiji is also helping out on this while he is visiting Caltech.
• The 40M FSS bandwidth was found to be too low. This was traced to somehow, again, using the wrong beam after double passing. We were using the correct +1 beam after a single pass but using the -2 after the double pass. We checked it by driving the VCO and making sure there was no wiggle on the double passed beam. This should get us back into shape.
• Work is ongoing with the image processing with GigE camera and feedback to the masses for beam centering. ezca libraries have been incorporated and now the thing is getting some testing.
• 166 MHz sidebands are getting tuned around in frequency to see if we can get better locking onto the +166 MHz SB instead of the anti-spring -166.

July 18, 2008

• Work continues on the project to measure the absolute length of the arm cavities with an external laser. There is lots of alignment drift in the external setup.
• The 40M PMC has degraded greatly over the years. The transmission is low, probably from contamination and in the last few months it seems as if the PZT has slowly started to fail and now the PMC loses lock every hour or so. Calibrated cavity sweeps show that the PZT response in m/V is down by a factor of ~5 from spec. We only get ~1 micron / 400 V. We are now on the prowl for a spare PMC...
• Calibrated the VCO and thereby the frequency noise monitor (IOO-MC_F) of the PSL to ~1%. There is a significant harmonic distortion seen when looking at the drive to the AOM via a 20dB directional coupler (http://dziban.ligo.caltech.edu:40/40m/687). Not sure if this is 'normal' or not -- we will try to clean this up with a bandpass filter.
• Whitening filters being designed and constructed to interface the fiber noise experiment to the 110B ADCs.
• Exciting developments in the DR IFO Locking after much late night work from Rob, John, and Yoichi:
  • DR IFO locked using the gradual method described last week.
  • Length offset introduced into DARM as done before in the power recycling configuration and done at Livingston for DC readout locking.
  • This is, we believe, the first demonstration of DC Readout on a dual-recycled Fabry-perot interferometer.
  • This was all done with the DR coupled-cavity locked on the -166 MHz RF SB. This makes the phase shift such that the longitudinal optical spring in DARM is actually an anti-spring, not just an unstable spring.
  • Lock Acquisition is much harder in the good spring (+166 MHz) configuration (we want the good spring  configuration for the NS/NS optimized AdvLIGO interferometer).
  • We suspected that there might be an accidental higher order transverse mode resonance of the +166 MHz SB in one of the arm cavities. J Miller calculated where all the HOMs show up and it seems that by bad luck the TEM01/10 modes of the +166 MHz SB partially resonate in the arms (down by a factor of ~40 in field from the carrier). WE are going to measure the g-factor of the arms and try changing the modulation frequency by a couple kHz to avoid this bad place. If that's not enough we may vent to shorten the MC so that we can increase the 166 MHz frequency by ~100 kHz.

July 11, 2008

• Full lock of the 40m DR IFO (first time in a long time): elog
  • short, Michelson DOFs are locked using the double demod signals
  • arms are locked using a combination of DC power and RF signals
  • details are in the graphic in the elog link
• Fiber noise experiment setup ongoing: setup of a Mach Zender to measure the phase noise in various kinds of 50m long fibers.
• Injection of an external laser to make high precision measurements of the macroscopic cavity lengths: Laser aligned and beat signals acquired. Hopefully first length measurements will be gotten next week. Koji Arai, who is working on this with Alberto, is off at LHO until the 21st.
• GigE Cameras look good. Image acquisition, streaming video, server/client code working. Looks like we will purchase several of these CMOS cameras from Prosilica for the sites in the next couple weeks.
• New 40m gateway machine is being set up as the CA, conlog, SVN, elog, etc. server for the 40M. This will also serve as the mDV and possibly the NoiseBudget code repository for the sites.
• Eric, Sharon, Jenne, Masha and Max helped Bob cleaning the large baffles.
• Yoichi looked at MZ board to understand the weird behavior. Converting the controls directories to svn. Set up http access to the svn on nodus. Moved a GigE camera to the PMC trans. port. The refresh rate was too slow when viewing the picture through ssh. With the help from Joe, installed the viewer application on one of the lab. laptop. Participating the lock acquisition effort (almost) every night.

June 27, 2008

• Lock acquisition: Rob, Rana, and John continue to work on reducing the CARM offset using a DC signal (SPOB DC). They were able to get up to arm powers of around 30 (where a single arm cavity lock is a power of 1) before instability set in and we would lose lock for, as yet, unknown reasons. They are trying different techniques to get around the problem. John tried using REFL_DC as the error signal for CARM rather than PO_DC, but found that it didn't work any better so he went back to PO_DC.
• Auxiliary laser locking: Masha and John are setting up a laser and Mach Zehnder to study fiber phase noise. Works as expected without the fiber!
• PSL: Yoichi found problems with the frequency stabilization servo controls, and traced it to a broken op amp on the FSS servo interface board. He replaced it and fixed the problem.
• Mode cleaner: Jenne, Rob and Yoichi were having trouble locking the mode cleaner, and traced it to a flaky cable in the MC servo box. They replaced it and now the MC is back to locking routinely.
• Absolute cavity length measurement: Alberto made cavity length measurements using cavity sweeping. As expected, the result inevitably has an ambiguity depending on which resonance do we take as an upper sideband. In order to resolve this ambiguity Steve and Koji performed a primitive non-optical measurement using a tape and photos, obtaining 38.48+-0.03 m and 38.67+-/0.03 m for the x and y arms, respectively. Koji is setting up a system for measuring the absolute lengths much more accurately, using precise scans of the FSR frequencies. In preparation for installing the system on our AP table, he made a nice annotated picture of all the beam lines on the table.
• Noise budget: Max continues work on his port of the noise budget code to the 40m. John is estimating contributions to the MC length noise.
• Adaptive noise cancellation: Sharon is learning about different methods for applying adaptive filters to improve the Mode Cleaner lock. She is exploring parameters for our adaptive FIR filter.
• Cameras: Joe an Eric set up a camera to view to the ETMX scattered light. Point scatters are clearly seen. Eric is writing beam profile analysis code in C to determine the beam position and width in x and y. IT works well for beams incident on the camera, but not so well for camera views of scattered light from in-vacuum optics. He is working on some signal processing techniques to clean up the high-frequency scatter noise.
• PEM: Rana revived the EPICS alarm handler, and adjusted the levels of several alarming channels to not alarm. He has instructed everyone to not ignore or turn off the system. It alerted him to excess particle count levels. He histogrammed the cumulative dust levels over the last year and set the yellow and red alarms at the 95th and 99th percentile levels.
• PEM: Max is working on resurrecting the Bartington magnetometer.
• Beam dumps: Steve continues to measure scattering off of wedged black glass and curved SS traps as a function of postion and angle.
• Safety: Steve led a lab safety walkthrough for Becky, Jenne, Max, Koji, Masha and Sharon.

June 20, 2008

• The SURF students have arrived!
  • Max Jones is working with Alberto on the 40m noise budget.
  • Masha Baryakhtar is working with John on auxiliary laser locking and fiber stabilization.
  • Eric Mintun is working with Joe on digital camera deployment and data analysis.
  • Sharon Rapoport is working with Jenne on noise suppression techniques.
  and Rana is the uber-mentor-of-all.
• Digital camera system: Joe is coding up client/server applications for display and processing of camera images.
• Koji Arai is visiting. He is working with Alberto on measuring the arm cavity lengths (to know how much RF leaks into them) and the recycling cavity lengths (to know the RF phase shifts in them).
• PSL: The PSL NPRO temperature is glitching; we suspect that the NPRO TEC temperature stabilization is broken. Steve and Yoichi did some heating tests; it looks like the LTEC loop is working but the readback is broken. Then they did a MOPA temperature scan, indicating that the temperature loop was working fine. They found a nice working point, and will keep an eye on it.
• PSL: Youichi is studying the PMC transmission (59%, quite low), and is working to optimize the alignment, mode matching, input beam quality, etc.
• Mach Zehnder: John has been investigating MZ glitches and railing.
• CO2 laser tests: Phil, with Steve and Rana, are setting up to test the eLIGO TCS laser in the east arm vacuum tube of the 40m. They are writing an SOP for the CO2 laser operation.
• Front end controls: all the front end control computers went down Wed night for unknown reasons. Everyone is getting good at bringing everything back up in the right order.
• Front end controls: Alberto, Rich and Sam installed Megatron (aka Rolf's Big Brain), the computer for the new control system being assembled by the CDS group. They measured the current drawn by this one computer: 8.1A!
• ITM upgrade: Rana is working with Garilyn on the polishing and coating specs.
• Locking: Rob, John and Youichi continue working on full DRFPMI lock.

June 13, 2008

• DRFPMI locking: Rob and John are making steady progress in lock acquisition of the full ifo. Things get twitchy and lock is lost as they ramp down the CARM offset and ramp up the power.
• Digital camera network: Joe is measuring the noise and response of another, higher resolution CMOS camera. The first CMOS camera (Prosilica GC750), when uniformly illuminated with scattered light, showed strange patterns (like PCBs); this new one doesn't.
• Noise cancellation: Caryn is working on filters that predict the mode cleaner fast and slow servo signals using data from accelerometers, seismometers, and microphones.
• Steve is seeing very little reflection or scattering off of his curved-wedge SS beam trap.
• PSL: Youichi is looking at the PSL transmittance.
• Youichi received 40m lab safety training from Steve.

June 6, 2008

• Upgrade: We met with Rolf and Jay to finalize plans for the 40m controls upgrade. We have "Rolf's big brain" in the lab, it is named "megatron".
• Tobin was visiting and he worked on the svn code repository and other computer-related work.
• Digital camera network: Andrey and Joe have been testing several cameras. They have been taking beam scans, and Andrey wrote beam profile gaussian fitting and displays in matlab. They're finding dead and hot pixels, taking flat-fields, etc.
• Steve is taking measurements of the scattering off his his wedged SS beam traps.

May 30, 2008

• Youichi Aso has arrived, and will be working on the new 40m upgrade. Please make him welcome!
• Caltech undergrad Caryn is developing MISO filters of different orders, with various sampling rates, for adaptive noise cancellation.
• The mode cleaner alignment has been a bit off since the computers went down last week. Rana worked on the alignment, and then John fiddled with the input periscope, to maximize transmission, minimize reflection, and get the beam back on to the WFSs.
• Rana has been working on improving the PMC mode matching and transmission.
• Joe and Andrey installed new GigE 24 port switches on our electronics racks, in preparation for rebuilding the whole ethernet network.
• Tobin is visiting. He set up an EPICS gateway server on our new gateway computer, and moved the NDS Proxy program onto it.
• Steve completed the installation, alignment, and centering of the new ETMY oplev after the old laser died.

May 23, 2008

• Oplevs: The ETMY oplev HeNe laser died. Steve replaced it and we were able to restore alignment and lock the arm cavities. Andrey realigned the MCR beam onto the WFSs and restarted those servo loops.
• Digital camera network: Joe continues to work on the network infrastructure.
• John Miller and Rob Ward returned from the GWADW in Elba.
• Tara Chalernsongsak, new grad student working with K. Libbrecht, was introduced to the basics of the 40m operations. Caltech SURF Eric Mintun received the 40m safety training. He'll be working with GigE cameras with Joe.

May 16, 2008

• Front end controls: Our front end controls network went down on Monday for unknown reasons (but EPICS controls are suspected). Andrey, Joe and Alex brought it all back and restored the SUS damping, oplevs, etc. On advice from Alex, they restarted the main EPICS control boxes, making it possible to restart the front end.
• Digital cameras: Joe continues to work on testing the digital cameras with PSL pickoff beams.
• Adaptive noise cancelling: Caltech undergrad Caryn is developing different adaptive filters to reduce MC length noise.

May 9, 2008

• Auxiliary laser locking: Steve is ordering 50 m of polarization-maintaining fiber, and Aidan has designed a system for measuring the phase noise of 1064-nm light down it and stabilizing it using an AOM, in preparation for using it to bring light from the PSL to an ETM for arm locking.
• Noise cancelling: Steve and Rana are purchasing a Guralp seismometer to sense seismic motion for adaptive noise cancelling.
• Frequency noise: Rana is tracking down a hump in the MC frequency noise spectrum. He suspects that it's due to some anomalous scattering on the PSL table, but it's not obvious where.
• Rana suspects that the relatively high power density on the ITMs could be producing thermal lensing; calculations in progress.
• Locking: Rob continues to work on DRFPMI lock acquisition, and has been switching some RF photodetectors around.
• Camera readout: Joe and Andrey are testing the readout of a GC750 GigeCam with some picked-off PSL light.

May 2, 2008

• Upgrade: Rana continues to work on the budget for the 40m upgrade, including new CDS controls, new ITMs, vacuum work, etc.
• MC WFS: Rana and John discovered that the MC WFS signals were not being properly whitened on their way to the ADC. Rana modified the whitening board to hardwire it on, and was able to increase the WFS gain to set the UGFs of the 4 loops to 4.01 Hz.
• Front End: Rana installed new FE filters for downsampling and upsampling.
• PEM: Andrey continues to improve the readout of the weather station and other PEM channels.
• DAQ: Alex's new data concentrator seems to be messing with the testpoint manager, so Rob killed the former.
• Scattering: Steve continues to make measurements of the scattering off of various beam dump materials.
• Backup: Alan fixed the ssh connection to the Powell Booth archive with help from Dan, and restarted the disk backup of the 40m controls software and trend frames.
• Network: Joe and Andrey surveyed all the networked computers in the lab and made a "network map", in preparation for upgrading the network to GigE.
• Computers: Joe commissioned a new gateway computer, nodus.
• Computers: The old vacuum controls computer has been replaced by linux3.

April 25, 2008

• Controls electronics upgrade: Jay Heefner presented plans by the CDS group for upgrading the 40m controls to the new AdvLIGO PCIe-based system. Rana is drawing up a budget for the 40m upgrade.
• Network: Joe continues to develop plans for upgrading the ethernet controls network for higher speeds. He is documenting the existing infrastructure. Andrey helped him trace and inventory existing cabling.
• Laser safety: Steve has been measuring the laser power on the PSL table and the AP table, and installing high power metal beam shields in various places.
• Dual-recycled locking: Rob and John continued to work on improving lock acquisition of the full AdvLIGO configuration. They continue to struggle with finding a good CARM signal for the common mode servo.
• Mode cleaner WFS system: Rana tried to diagonalize the MC WFS system and couldn't. Rob found a problem in the way the dewhitining switches were implemented. It apparently now works, and worked a little more on the MC WFS, making many changes to the sensing matrix script and also to the MCWFSanalyze.m script.
• Adaptive noise cancellation: Rana and Caltech undergrad Caryn Palatchi are working on developing effective and efficient FIR filtering.
• Weather monitoring: Andrey continues to work on improving the weather station performance and readout.
• IST Building construction: Heavy construction work continues to prevent working with the interferometer during construction work hours.

April 18, 2008

• Most of the week has been occupied by preparations for the 40m Technical Advisory Committee on Thursday, 4/17/08. We have prepared some slides on recent work (noise cancellation, aux lock acquisition techniques) but the main thrust was on plans foran upgrade to the 40m to more closely resemble the most recent design of the advLIGO optical configuration. Slides, and the latest draft of the upgrade technical document, are here: http://lhocds.ligo-wa.caltech.edu:8000/40m/TAC_Meeting-Apr_08
• Rana used Foton to make up some new filters which can be used all over the project in order to downsample/upsample.
• Physical Environment Monitoring: Andrey has fully integrated the weather station with our data acquisition system and documented trends for the last few weeks.
• IST Building construction: Heavy construction work has made it impossible to work with the interferometer during construction work hours.

April 11, 2008

• Upcoming TAC meeting: The 40m Technical Advisory Committee will meet next Thursday, 4/17/08, 8:30 Pacific. Ken is unavailable, so Valera will chair and prepare the TAC report. Slides will be posted on the 40m wiki http://lhocds.ligo-wa.caltech.edu:8000/40m/TACs a microsecond or two before the call starts.
• Rana and Stefan continue to develop the plans for upgrading the 40m for the new AdvLIGO broadband RSE control scheme, including: modifications to the ITMs (mass, transmission at 1064 and at 900 nm); different modulation frequencies; different controls signals for broadband RSE; longer recycling cavities; elimination of the Mach Zehnder; new front-end controls (BorkSpace); and other cool innovations.
• Modeling: Lisa continues to develop Optickle and Looptickle modes of the 40m in PRFPMI and DRFPMI configurations.
• GigE network: Joe has drawn up the current 10/100BT controls network and is making plans to upgrade it to GigE for readout of the digital cameras.
• Suspension controller problems: Much reduced, now that Rob has swapped some of the Penteks in the sus crate.
• Physical environment monitoring: Andrey and Steve have brought back the readback of the inside and outside weather station sensors.
• Beam dump development: Steve continues to measure reflectivity of various candidate materials.

April 4, 2008

• DR Locking: John continues to work on aux laser arm locking and 3f locking, and gave the group a status talk on Wednesday.
• GigE network: Joe is refining his plan for a new GigE network for readout of the digital cameras.
• Suspension controller problems: The suspension controller continue to time out, requiring reboot. Rob swapped various things in the c1susvme2 crate. Finally he swapped the two Penteks with ones from the asc crate, and the problem seems to have been reduced dramatically.
• DC readout PDs: Jamie Rollins was visiting, and he made some measurements of the non-linearity of the response of some mock-up DC readout photodetectors, and compared with a simple model of non-linearity. The test PD was pretty linear, with non-linearity parameter beta < -1.
• Physical environment monitoring: Andrey resurrected the weather station readout into EPICS, which has been broken for years.
• Beam dump development: Steve continues to measure reflectivity of various candidate materials.
• Rana gave tours of the 40m lab to ~40 visiting grad school admits.
• Safety inspection: We had a safety inspection of the 40m and bake oven labs last Friday.

March 28, 2008

• DR Locking: John was able to acquire lock of the DRM with 3f signals, and hold the lock, albiet rather shakily.
• GigE network: We need a new GigE network for readout of the digital cameras that Joe is working on, as well as to make the rebooting of front end processors lass flaky. Joe is beginning to draw up a plan. He's also taken some pics with two different cameras (GC650 (CCD) and GC750 (CMOS)) for comparison.
• Noise cancellation / adaptive filtering: Rana borrowed a Guralp seismometer from MIT, and he and the Abbotts have wired it into our PEM system. Alex routed the signals from the PEM to a BorkSpace processor (c1ass), Matt wrote Adaptive filter code ("Filtered-X-LMS") to generate control signals which will be routed to the unused SUS_MCL channels of all the optics suspension controllers (except MC2). Matt was able to produce adaptive filtering that can achieve results similar to those of the MISO FIR Wiener, potentially reducing seismic noise by factor of 10 or more in the 1-25 Hz frequency band. In the current setup, we have 6 accelerometers and 1 seismometer around the MC to be used to demonstrate the principle of the whole thing.
• Suspension controller problems: we continue to have problems with c1susvme2 getting out of sync and/or giving a red status light, requiring power cycling. c1susvme1 is also misbehaving; we suspect that the fan installed to help cool the crates is causing the problem.
• RFAM monitor: Valera visited the lab and measured the RFAM after the RF stabilization box in the 166 MHz modulation path.
• Squeezer goes home: Tim Bodiya came by, completely dismantled the squeezer from the PSL table, boxed it up and shipped it back to MIT. We now have a lot of empty real-estate on the PSL table.
• New PCIX-based controls: Rolf is beginning the design of new PCIX-based (AdvLIGO-like) controls to replace our VME-based system.
• Safety inspection preparation: Steve, Bob and alan prepared the lab, and prepared documents, in preparation for an upcoming safety inspection.
• Vacuum: small turbopump TP2 (forepump for main TP1) failed and was replaced by Steve. The software interlock (which closes the main valve and shutters the laser) worked fine, which confused the people working on the IFO at the time.

March 14, 2008

• DR Locking: Rob continues to make progress with locking the full inteferometer in the AdvLIGO configuration. Using single-demod signals (carrier - RFsidebands), the DRMI (MICH, PRC, SRC) locks quickly, followed by one arm, then the other (slightly off resonance). The whole process takes a few minutes. The SRC controls are handed off to double-demod signals (just RF sidebands). The handoff to CMservo analog path is still problematic; amongst other things, the MC suspension bounce modes at 16 Hz produce too much noise. Work continues.
• Aux Beam Arm Locking: John continues to work on this, not much progress this week.
• GigE Cameras & Image Processing: Joe is testing and comparing two different cameras; one is CMOS, the other is a CCD. Joe continues to assemble a dedicated GigE backbone to test network capabilities as regards frame rate, resolution, etc.
• Adaptive filter noise cancellation: Matt Evans is working with Alex to set up a Borkspace system, and Alex has installed software to route PEM signals to the c1ass computer for processing and thence to the SUS_MCL for feedforward. Rana is looking at purchasing more PEM equipment: Guralp seismometers, microphones, magnetometers. Josh is looking at PEM signals suitable for this (SNR > 10).
• 3f locking: Lisa, Rana, Rob and John have implemented a system for controlling MICH and PRC using RF demod signals at 3f = 99 MHz the REFL port (REFL99_I and REFL99_Q). These signals come primarily from the sidebands resonant in the PRC, not the carrier, so it should be less sensitive to what happens when the arms lock and ramp up. They made minimal hardware changes, modifying the 199MHz demod board, changing the local oscillator to 99.58 MHz. After tuning the demod phases and removing offsets, they locked both the PRM and the DRM in a stable way. They weren't able to acquire the lock in the DRM configuration directly by using the 3f signals; they needed instead to use the single-demod signals first, and switch to the 3f signals once the lock was already acquired. They are working on improving the signals so that they can be used directly for locking. Meanwhile, Lisa has returned to MIT - come back soon!
• Steve continues to work on simple beam dumps. He is measuring BRDFs, relative to white paper.

March 7, 2008

• DR Locking: Rob is making progress with locking just the DRMI. Acquisition with the single demod signals is working well, and handoffs to all double demod signals using the input matrix ramping worked several times with the scripts. Arm locking (without the DRMI) also works fine. Up next will be more work to put together the DRM+ARMs.
• Aux Beam Arm Locking: John continues to work on this, with help from visitor Lisa Barsotti. They are working on getting clean signals through the PDs, cleaning up RFAM contamination, and looking for excess frequency noise. John is measuring the RFPD responses.
• GigE Cameras & Image Processing: Joe continues to assemble a dedicated GigE backbone to test network capabilities as regards frame rate, resolution, etc.
• Noise cancellation: visitor Matt Evans is looking into setting up a Borkspace system for noise cancellation at the 40m.
• Youichi Aso is studying the AdvLIGO and 40m modelling that has been done with Optickle.

February 29, 2008

• DR Locking: Continued problems with switching from single to double demod with just the DRM. Hopefully, the new RFAM monitor will shed light on this.
• Aux Beam Arm Locking: More RFAM. The fringe gotten from locking from the backside is very weak and can be totally swamped by a time varying RFAM. We now suspect the fiber of converting PM to AM. Next move will be to move the EOM to the ETM side of the fiber. There's also an analysis done by Lisa Barsotti (not on the Wiki !!) on this type of locking.
• Beam Dumps: We have measurements of BRDF from mirror polished steel and copper for making of a low scatter, high power dump. Key in getting low scatter with metals is to align the grain of the polish parallel to the incident beam direction. A preliminary mockup of the steel dump looks near as good as a black glass wedge (to the limits of our scatterometer setup). Plots to follow in the elog as soon as Steve gets back from skiing.
• GigE Cameras & Image Processing: We now have 1 CMOS and 1 CCD camera. We're putting them into a dedicated GigE backbone to test network capabilities as regards frame rate, resolution, etc. We are surveying vendors and will purchase from 1 more before getting back to software development (fitting, centroid tracking, diff detection, etc.). We are interested to see if network hardware will let us get rid of the analog MUX's and go to a completely GigE camera solution for the sites.

February 22, 2008

• Rob gave a LIGO seminar on beyond-Advanced-LIGO technologies, concentrating on new ideas for prototyping at the 40m. His slides are here.
• Alberto has assembled his RFAM monitor circuitry (StochMon) and is connecting it up to EPICS channels, with help from Ben. In the course of that work, the laser accidentally tripped off. It took a few hours to recover and come back up to full power, but it's ok now.
• Alberto goes back to Italy for a few months, this weekend.
• John continues to work on aux laser arm locking, but is running into some serious problems. Work in progress; a more complete report next week.
• Joe is researching and placing orders for new GigE cameras, switches and PCI network card.
• Stefan continues to work on developing an AdvLIGO-like model for 40m prototyping.
• We've been talking to Caltech student prospective SURFs.
• The Tuesday 5.0 earthquake in northern Baja California tripped off our suspension controllers, more than once, but they came back up fine after things settled down in Mexico.

February 15, 2008

• John and Sam are continuing development of auxiliary laser arm locking. They have light from the arms, exiting ETMY, coupled into a long fiber. They are setting up RF signals, EOMs, demodulators for PDH signals, etc. They estimate that such a system would have a shot noise sensitivity of a few 1e-15 m/rtHz with the undercoupled arm cavity.
• Rob has been working on dual-recycled FPMI locking. There were a variety of problems with double-demod signal handoffs not working well, so Rob is looking for more robust signals.
• Rob noticed that one of our suspension controller cpus (susvme2) is having bursts of "un-syncy lateness", potentially interfering with locking activities. This behaviour appears to be getting steadily worse. To be debugged.
• Rob is preparing a LIGO seminar for next Tuesday on "Thinking Beyond Advanced LIGO".
• Alberto has built a prototype RFAM amplifier, and is debugging it.
• Go's paper on injecting squeezed vacuum into the 40m has been accepted for publication in Nature Physics. The editors are hinting that they might put some relevant artwork on the front cover, so Go and colleagues are thinking of something appropriate.
• Sam has cloned our DC readout EPICS screens at LLO, eliciting comments about the color scheme.
• Joe is finishing up current GiGE camera code implementations. On Rana's urging, he is accelerating his develoment of a digital camera network for IFO diagnostics. He will start testing different camera models and building up a gigabit network infrastructure.
• Steve installed a new RGA head a few days ago and it seems to be working fine.
• Most of the heavy ground prep work for the new IST building is done; they will start pouring concrete for the foundation soon.
• A mag 5.1 EQ in Baja California tripped off all our suspension controllers this weekend.

February 8, 2008

• Joe is working on GiGE camera interface, which can now be run on all linux based machines in the 40m. Working on getting capture rates up (currently can sample a beam and save the images at about 12 Hz). Also working on some basic matlab functions to analyze the beam images. Camera has currently replaced the PMC transmission video camera.
• Stefan has a "Bork-space" digital control system working in AdhikariLab, and has used it to lock the PMC to the 35W laser (operating at low power, so far).
• Alberto is waiting for parts to assemble his RFAM box, which filters four different RF frequencies.
• Steve and Bob are baking out our repaired RGA head and will be checking it out this week.
• John now has optical fibers running from a laser on the SP table to the end of the y-arm, and has coupled it in to YARM through ETMY. He has coupled ~75% of the light transmitted from the y arm, through the fibre, back to the SP table. He's still looking for FP flashes, and is setting up a PDH system for aux arm locking.
• Rob continues to work on full interferometer locking. He's debugging double-demod signals, and is also exploring other (differential-demod) signals for locking the central part.
• David has shipped two tip-tilt mirror suspensions to LLO for the DC readout chain.
• Phil Ehrens has set up an SVN archive as a versioning repository for the 40m code base. Alan is populating and debugging it.
• Construction of the IST building continues to generate large amounts of vibration noise at the 40m during the day, but we are surviving. Rana's BLRMS monitors are providing very useful information.
• Rana met with representatives of a new geology faculty member who wants to set up a large system in CES (the big building that the 40m wraps around) to study flows of rocks, sand, etc. Just the kind of thing we like to have next door.

January 31, 2008

• Tobin is off to LLO to begin his LSC Fellowship.
• Joe has succeeded in reading out the GigE camera and getting the data into Matlab for display and manipulation. He will now set up a system for reading out multiple cameras, and compressing and archiving the beam spot images on a regular schedule. Next is to automatically destermine the spot centroi, width, and non-gaussianity, and log these data to frames. Then, deploy many such cameras throughout the interferometer.
• Alberto has a final design for RF filters to isolate and monitor the power transmitted through the mode cleaner at all RF frequencies of interest: 33, 133, 166, 199 MHz.
• John continues to work towards locking the arms with auxiliary lasers. He has run an optical fibre from the SP table to the ISCT at the Y-end. In the coming days he will try to mode match the beam from this fibre into the arm through ETMY.
• Rana wrote a matlab script to produce band limited RMS trends from our accelerometers. It mimics the code written by Ed Daw which makes the seismic FOMs at the sites. It pulls data from mDV and puts the results in EPICS channels.
• Andrey continues to study the effect of damping gain of ITMX and ETMX on the BLRMS noise of XARM. He is seeing a broad minimum of noise for some values of the damping gain (Q). We is now working on a noise model to understand this better.
• Rana continues to work on getting meaningful data from the accelerometers and seismometers, to be used for adaptive filter noise cancellation.
• It has been established that, as a general rule, "clicking random blue buttons chaotically" is not a good problem solving technique. It is thus now explicitly discouraged as an option in the LIGO 40m Lab.

January 25, 2008

• The Wednesday night rains caused a flood of water and mud in the 40m control room and IFO hall vertex area. Quick hard work by Andrey, Alberto, John, Steve and Rana staved off any major equipment damage. The next-door construction crew had blocked up the stom drains, causing a lake to form at the 40m north wall, and water seeped right through the wall. The Caltech facilities guy came and unblocked the drains. bu Thursday noon, Steve and the janitor had cleaned up the floors and electronics racks, and the crew is starting to bring things up. Looks like there's no major damage.
• Rana conceded a bet yesterday, and treated the 40m team to sugar napoleons.
• Rob, John and Andrey are making great progress in locking. After overcoming several obstacles, they got the interferometer fully locked in a power-recycled FPMI state, and measured a noise spectrum that is not much different from what we had months ago; it bottoms out at ~ 2e-18 m/rtHz in the 1-3 kHz range. There's more noise at 100 Hz, but that is understood and can be reduced.
• Then Rob did some dual-recycled Michelson locking. He's working on a bunch of issues, including the handoff to the double-demod signals.
• Andrey has succeeded in measuring the suspension Q's as a function of damping gain, getting good results. He continues to work on minimizing the XARM noise by optimizing the damping gain of the TM optics.
• Our RGA controller is back from repair and Steve will re-install and test it out, soon.
• Alberto continues to perfect his RF filters for the MC transmitted RFAM monitor. He's designing low-pass filters to make the LC resonant circuits more symmetric, to achieve -40 dB of attenuation at +-33 MHz around each filter (33, 133, 166, and 199 MHz).
• Tobin and John are testing a long fiber to see if it can bring a PSL pickoff beam to the ETMs for auxiliary laser arm locking, with acceptable phase noise. In doing so they used our GigE camera for a physics application for perhaps the first time, viewing the transmitted mode from the fiber during initial alignment. Meanwhile, Sam is doing a similar fiber test at LHO.
• Kentaro is here. He's using Optickle to study the response of GEO600, and is trying to understand why they don't see an opticle spring at 10 Hz where they expect it. Rana suspects the problem is due to the difficulty in understanding the triple-pendulum suspensions (and aux loops) at 10 Hz. Kentaro is also working on the design of a 10-meter SQL interferometer at Hannover.

January 18, 2008

• Rob and company continue to make good progress in lock acquisition. Rob is concentrating on the power recycled FP Michelson configuration, controlled by DC readout after the OMC. He got through the handoff of DARM to the OMC, and is now working on handing off CARM to REFL166.
• Plans are developing for prototyping a auxilliary laser arm-locking scheme, where some light picked off from the PSL is fiber-fed into the ETMs. Tobin and John are setting up a system to measure the phase noise of a long optical fiber. Rana is talking with Jun Ye of JILA, who has developed phase-noise cancelling long fibers.
• Alberto has finalized the schematic for the RF monitor board based on passive Chebyshev band-pass filters. With help from Rich and Ben, he designed and built a test circuit for the 166MHz line using tunable inductors. He gets more than 20 dB of isolation after 33MHz. He's working on a design with more filter orders for better rejection.
• Pinkesh and now Joe B have gotten the GigE camera to work, and Joe is now beginning to develop software for automated readout and analysis of beam profiles.
• Andrey continues to work on suspension controller gain optimization for the arm cavities.
• The ETMY suspension controller has been tripping off lately. It appears to be due to large seismic events. Andrey will move an accelerometer over there to monitor it.
• The signal from the BS UL sensor continues to be ratty, and we believe it is due to scattered light: either on the OSEM photodiode itself or the wires carrying the signal out. Realignment doesn't help much. For now, the UL signal has been switched off in software.
• We have assembled a list of potential SURF08 projects at the 40m, including: auomated beam scanning and aligning; adaptive filtering for noise cancellation; development of our PEM system (weather, voltage monitors, microphones, etc); commissioning our automated dither alignment system; noise budget work; further study and suppression of optical lever noise; Mach Zehnder improvements; implementation of a new RF modulation scheme for length control; aux laser arm locking; SPI stuff.
• IST building construction has begun in earnest with big heavy ground breaking machinery. There can be a huge amount of seismic noise during the day.
• We have been informed that there are plans for a geophysics lab in the CES building (the one we're wrapped around), including a setup that can best be described as a "giant rock tumbler". We have made it clear that vibrations cannot be tolerated after hours, when we try to operate the interferometer.
• Next week is Tobin's last before moving to LLO. Beer at the Ath, Wednesday ~ 5pm!

January 11, 2008

• John, Tobin and Rana have been making good progress in lock acquisition. They were able to lock the DRFPMI (i.e., full IFO) a couple times last night.
• Rana and others are thinking about using auxilliary lasers (or a fiber-run pickoff of our main laser) sent through the ETMs to lock the arms independently of the DRMI.
• The PSL, FSS, PMC and ISS are all behaving well, except that the PMC servo sometimes rails and requires relocking. To be investigated.
• The mode cleaner and most optic suspensions are behaving, except ETMY watchdogs seem to trip a lot and the BS UL OSEM has a ratty signal; seem to be picking up stray 1064 light.
• Alberto continues to work on the RFAM monitor photodetector design. He's now working on the prototype PCB.
• David continues to work on mode matching for the 35W laser.
• Rob is working on his thesis.
• Pinkesh and Tobin continue to work on reading out the GigE camera for use as a digital beamscanning device.
• Steve sent out the RGA for repair and they say it's working and are sending it back.
• Electricians have now installed a row of ethernet ports in our office and control rooms.

December 21, 2007

• Rob submitted a final draft of the DC Readout Amaldi paper with responses to referee's comments.
• Tobin continues to study the PSL RIN with the instensity stabilization servo on and off.
• Alberto continues to test different designs for his broad-band MC RFAM monitor circuit.
• Andrey continues to optimize the suspension damping gains of ITMX and ITMY by minimizing the XARM rms noise.
• Steve and Tobin have discovered that a particular kind of laser safety glasses that we have been using for 2 years at the 40m (LaserVision glasses with uvex T93 Filter reflective dielectric coatings for IR) are defective. They have very good visibility in comparison with our absorptive glasses, so we have been favoring them. But the dielectric coating degrades (for unknown reasons), thereby providing no protection from laser light. We have collected all 6 of our glasses and gave them to Bill Tyler for disposal. We have notified all site safety officers at CIT, MIT, LHO, LLO, GEO, and ACIGA.
• Steve has measured the IR transmission of all of our other (absorptive) glasses; they were all unmeasurable with a 170mW Crystal Laser 1064nm laser.

December 14, 2007

• Rob is preparing a response to the referee comments on his DC Readout Amaldi paper.
• Tobin wrote a script to automatically measure the in-loop and out-of-loop RIN of the PSL instensity stability, with the servo on and off.
• Tobin is working on resuscitating the readout of the 40m weather station, which hasn't worked in years.
• Alberto is designing and testing filters for the broad-band MC RFAM monitor circuit.
• Tobin fixed the Mach Zehnder autolocking script so that multiple copies can't run.

December 7, 2007

• Rana compared the MC_F spectra between H1 and the 40m, and found that our frequency noise is about the same as H1 above 200 Hz, except for the enormous power line spikes. Our acoustics are better and the noise above 1 kHz levels off at the same flat floor (the phase noise of the VCO) as H1. So he's declaring the FSS to be just fine.
• Tobin continues to work on the PSL ISS servo to fix noisy preamps, reduce the noise, and increase the gain. He replaced the short PD cables with longer ones to make it easier to work on the system. He took some RIN spectra and found significantly increased dark noise when the cables and wiring were changed. At higher resolution, nearly all of the excess dark noise on the ISS is in 60 Hz harmonics. With the loop closed the noise floor is 10^-7 RIN or better from 40 Hz to 8 kHz. He's hunting for the source of the increased 60 Hz coupling.
• Rana and Rob are developing all kinds of useful data monitoring functions in matlab (BLRMS seismic, line monitoring, MCT RFAM, etc), inserting computed channels back into EPICS for DAQ logging, and stripchart recording.
• Andrey has securely mounted all six accelerometers to the floor by ITMX and ETMX, and is measuring seismic spectra and coherences.
• Andrey replaced the dead ETMX oplev laser.
• Alberto is designing a board to measure the power of the five RF lines on the mode cleaner transmitted (MCT) light. First he's trying RF filters to isolate each RF line. If that doesn't work well enough, he'll try mixers.
• Bob and Larry are putting in more permanent Cat-6 ethernet cables in the 40m office area, to improve the reliability of our GC network.
• Our weather station channels haven't worked for years; we're looking for someone to put in a new, better system (eg, like LHO's).
• David used the beamscanner to measure the profile, waist, rayleigh range of a test laser, in preparation for doing the same with the 35 watt eLIGO laser.
• Steve sent out the broken rga controller for repair.

November 29, 2007

• Rob and John are back to working on locking the full DRFPMI configuration. Better frequency stabilization is required, so there's been a lot of attention paid to the PSL servos.
• Rob, Rana and Tobin have been working on the PSL frequency stability servo (FSS), which is performing much worse than in the past. They have found and fixed many small problems. After this work, they got the UGF up from 100 kHz to ~250 kHz. Work continues.
• Tobin has been working on the PSL ISS servo to reduce the noise and increase the gain; he got the UGF up to 40 kHz.
• John measured the PSL pre mode cleaner pole frequency to be 380kHz +/- 59kHz. This corresponds to a finesse of ~936. The nominal design pole is at 488kHz and the finesse is 732.
• Andrey is working on modeling the transfer function from floor, through stacks and pendulums, to the test masses.
• Rob is developing the looptickle model of the current 40m, to add in oscillator noise couplings.
• Tobin installed "labca", which allows direct access to EPICS channels from within Matlab.
• Andrey continues to develop his model of the seismic noise at the test masses, including the transfer function from floor, through stacks and pendulums, to the test masses. He gave a talk at our 40m meeting, and created a page in the 40m Wiki.
• The vacuum system RGA controller is broken.
• Mohana brought a group of Boy scouts and their parents to see the 40m on Friday Nov 16.
• The system disk for our main nfs server (linux1) started dying, so Tobin and Steve got a replacement. Alex put the new hard disk into linux1 along with a fresh install of linux (CentOS).

November 16, 2007

• Construction has now begun on the new IST building. There is no access to the lab from the north. Jackhammers are banging away right outside the control room.
• Tobin and John have been working on improving the PSL Intensity Stabilization Servo (ISS). They probed the servo board and found lots of oscillations. The biggest was at 4 MHz, beats between the 33 MHz and 29 MHz RF sidebands on the light. John built a notch filter and installed it on the board. Rob measured the transfer function and RIN. They are now going through a list of further improvements as suggested by Stefan, including finding and eliminating other oscillations, replacing op amps, checking compensation capacitors, gain roll-off HPFs, making sure all differential signals really are differential, etc.
• John is also working on measuring the PSL PMC pole frequency.
• Rob and Rana spent some time diagnosing the degredation in the PSL frequency stabilization servo (FSS) UGF. There's much less light getting to the reference cavity than there should be / used to be. They aligned various things and got the transmission up 15%, but probably also need better mode matching, AOM throughput adjustment, etc. Work is ongoing, and is necessary to get the noise down sufficiently to lock the full interferometer.
• Alberto is getting involved in 40m activities. He'll be working with Caltech undergrad Max on implementing the new noise budget software at the 40m (and updating the code in CVS so that it runs at all sites).
• Alberto is also beginning the design of an RFAM monitor at the mode cleaner transmitted port, using a broadband RFPD to simultaneously monitor all relevant frequencies (33, 166, 133, 199 and 29 MHz).
• Andrey is working on modeling the transfer function from floor, through stacks and pendulums, to the test masses.
• Steve is designing a "wood's horn" copper beam dump for use with the eLIGO 35 W laser.
• Pinkesh and Norna have been working on the OMC suspension. With help from Chub, they reduced the stiffness of the wiring sufficiently to see a pretty decent vertical transfer function. Finally, they removed ALL the wiring and saw a vertical transfer function that is close to design. They are revising the layout of the wiring to minimize mechanical stiffness on the suspended OMC.
• Rana installed the new matlab-based ligoDV (previously GEO DV) to look at 40m data.

November 9, 2007

• The 40m Technical Advisory Committee (TAC) met today. Rana showed these slides, which includes results on suspension optimization by SURF Sonia Buckley, noise cancellation adaptive filtering by SURFs Keenan Pepper and Elena Gaspari, DC readout noise transfer functions from Rob, results from our MC drag-wiping exercise (see below), and recent results from Stefan and Rana on modeling the 40m length sensing and control system and noise sources (Bench).
• After completion of the drag wiping of the mode cleaner optics last week, Steve and company sealed up the MC chambers, and he and Andrey pumped back down on Monday. The pumpdown went smoothly (and slowly, to avoid stirring up dust).
• After returning to vacuum, Rob and company got the mode cleaner locked with little difficulty. The beam went through the Faraday isolator into the main interferometer, and the XARM was also locked with little difficulty. The MC was measured to by 2.7mm shorter than before the vent (not surprisingly), and the RF frequencies were all adjusted accordningly.
• Tobin re-measured the MC cavity ringdowns and Rob re-measured the MC pole frequency. Both measurements were approximately unchanged with respect to the measurements made before drag-wiping. We conclude that the inferred ~100 ppm loss per mirror is not due to dust particles on the mirror surfaces. This, plus some pencil beam scatterometer measurements made by Liyuan, suggests that the anomalous losses are due to many small coating defects.
• Tobin and Rob are optimizing the PSL intensity stabilization servo (ISS). They cranked the gain up to get a 35kHZ UGF without saturating the current shunt actuator. They are finding problems with the electronics chain, so there is room for much improvement.
• Sam and Pinkesh continue to make progress in commissioning the eLIGO OMC. They made a whole mess of characterizations of the PZT control and then locked the OMC with a PZT dither lock and a 600 Hz loop. They also worked on measuring the transfer functions of the 6 degrees of freedom in the OMS SUS. It will be disassembled next week for vacuum prepping.
• Tobin hooked our Tektronix scopes up to the network, gave them IP addresses, and wrote a script to access their data and get it on to the controls computers. Neat!

November 2, 2007

• Final papers from 2007 SURF students:
  • Sonia Buckley, U Dublin, LIGO-T070168-00-I, Automated Optimization of the Local Damping and Control for the Suspended Optics at the 40 m.
  • Joshua Weiner, Caltech, LIGO-T070266-00-I, Automated Analysis of the Transverse Modes of a Laser Beam Using a Non-Degenerate Fabry-Perot Cavity.
  • Alice Z. Wang, Georgia Tech, LIGO-T070193-00-I, Next-generation Radio Frequency Photodiodes for Gravitational-Wave Interferometers.
• In preparation for drag-wiping the mode cleaner optics at the sites, Rana wanted to test the procedures at the 40m. In preparation for that, Rana measured the MC transmission and took ringdown measurements, he and Tobin measured the MC optic resonances in the 28 kHz region, and Rob measured the cavity pole. Attempts to quantitatively measure the scattered light proved to be difficult. All measurements pointed to a finesse of 1460 and a total round trip loss of about 300ppm. We will see if this changes after drag wiping.
• Tobin and Steve perfected a drag-wiping method by practicing on the spare suspended small optic. Using a 50 microliter Hamilton syringe to deliver 16 microliters of methanol seems perfect for drag-wiping the small optics.
• Steve, David and Andrey began the vent on Wednesday morning. It went smoothly. The chambers housing the MC optics were opened. Rana, marked MC1's location with dogs and then slid the suspension horizontally to the table edge for easy drag-wiping access. The optic was thoroughly hosed-down with the dionizer. Drag-wiping commenced with Rob squirting (using the 50 microliter syringe) and Tobin dragging (using half-sheets of Kodak lens tissue). We drag-wiped the optic many (~10) times, concentrating on the center but also chasing around various particles and a smudge on the periphery. There remains one tiny speck at about the 7:30 position, outside of the resonant spot area, that we could not dislodge with three wipes. We drag-wiped MC3 only three times, all downward drags through the optic center, with Steve squirting and Tobin dragging. Some particles are still visible around the periphery, and there appears to be a small fiber lodged near the optic center on the reverse face. Andrey and Steve have opened up MC2 in preparation for drag-wiping that optic.
• We expect to pump back down on Friday.
• Sam and Pinkesh made great progress on the eLIGO output mode cleaner (OMC). They hung the OMC on its suspension, discovering many problems and potential solutions along the way. They mounted the four QPDs, got them into the digital system, and measured their response with and without light. After some effort, they were able to lock the OMC to the laser using the new digital controls.
• Our Nikon D40 camera has been missing, and some mysterious stranger is asking for a ransom of a plate of goulash.
• The SoCal fires have increased particle counts in the lab. All the HEPA filters are turned up to 100%.
• A largish (M5.6) earthquake in San Francisco on Tuesday sent our optics swinging.

October 26, 2007

• Andrey continues his work on optimizing the suspension damping gains for ITMX and ETMX to minimize the rms error on the XARM locking. He is focusing on conditions during the evening, when ground motion is lower.
• Rana continues the work on adaptive filtering of accelerometer signals to reduce the rms motion of the suspended optics, begun by Keenan Pepper and Elena Gaspari. He plans to purchase more Wilcoxon accelerometers and find permanent locations for them near the 40m chambers.
• Rana has installed Martin Hewitson's GEO dataviewer software (MATLAB-based) at the 40m. It now works with the LIGO NDS system with some NDS advice from Rolf (who's over in Germany this week); so now, GEODV is LIGODV. It's very cool.
• Rana proposes that we vent next week to drag-wipe the mode cleaner optics, as a test of the procedure being planned for H1 and L1. The MC's at the sites currently lose ~30% of the power and so if we can get it all back that's ~5 Mpc. The procedure will be to mark the suspension tower positions, remove the towers, drag wipe the suspended optics and then return to the stacks. We will see whether the MC transmission, finesse, losses, scattered light, etc, change after drag-wiping.
• The c1iscex computer, which controls the ETMX suspension, has lost timing synch, requiring reboot, very often. Rob found and replaced a ragged timing cable last week, and there hasn't been a problem with the timing synch since then.
• Sam, Pinkesh and company locked the eLIGO OMC using the PZT length actuator, and measured the PZT response. There are many resonances in the 4.5 - 10 kHz region. Pinkesh measured and calibrated the PZT transfer function. The two DCPDs were mounted to the glass breadboard, as well as the QPDs. The DC gain between the two DCPDs is, at first glance, different by 15%. They then integrated the cabling harness with the OMC. The OMC is now back at Bridge lab and will soon be mounted on the OMC suspension for further tests.
• Steve is monitoring the particle count and environmental conditions in the lab, while Southern California is on fire.
• Bob's bake oven lab will be quite busy throughout November vacuum-prepping eLIGO SEI HAM parts. Bob's new 16" oven is almost ready to go into operation.
• Steve noticed that the PSL chiller temperature went up by ~ 1.5deg C; he traced the problem to a piece of paper patially blocking the chiller's air intake.
• We have moved to a new ELOG which seems to work ok.
• Grad students Pinkesh Patel and David Yeaton-Massey got 40m lab safety training from Steve.

October 19, 2007

• Rob has submitted a draft of his Amaldi paper on "DC Readout Experiment at the Caltech 40m Prototype Interferometer" for LSC review. His talk at Amaldi earned him a James B. Hartle Award, and was chosen as a highlight paper for CQG.
• We had a 40m Tecnical Advisory Committee (TAC) meeting scheduled for last week, but because of travels, etc, it has been postponed till Nov 8, 8:30 AM Pacific.
• Stefan has used loopTickle/Opticle to build a length sensing and control model and noise budget for AdvLIGO and a next incarnation of the 40m, featuring the new 9.4 / 47 MHz control scheme and 10g ITMs (to move the optical spring peak above 200 Hz).
• Andrey has been working on optimizing the suspension damping of ITMX and ETMX, by locking the XARM and adjusting the damping gains to minimize the rms noise on XARM_CTRL in the region of the suspension resonances, while keeping the rms noise in the stack resonance region under control. He is documenting the XARM lock procedure in the 40m Wiki.
• In the process, Andrey is exercising the new matlab-based dataviewer and dtt facilities develooped by Justin.
• Rob continues to work on issues related to locking the full IFO in dual-recycled-FPMI configuration. He'll measure the laser and RF oscillator noise couplings to RF readout and DC readout in that configuration, to compare with his measurements in the PRFPMI configuration and with his models. He notes that there are currently some large offsets on small signals in the REFL port, not yet understood.
• There was a mag 4.2 quake in Wrightwood on tuesday early morning, which got all our suspensions shaking. This gave us a nice opportunity to study the ringdowns of the suspended optics, and has uncovered problems with one or two OSEMS (MC2 side is not working).
• We continue to have problems with c1iscex, which controls the ETMX suspended optic. It times out and requires rebooting, every few days.
• On Mon Oct 8, 5:45 AM where was a short power outage. Steve and Rob went through all the coldstart procedures and everything came back up OK.
• Rana wants us to all start using a new improved 40m elog.
• Sam, Pinkesh and company are measuring the optical properties of the Enhanced LIGO OMC at the 40m clean room. They measured the finesse, transmission, and modal structure, and estimated the round-trip losses. It will be moved to Bridge for integration onto its suspension sometime soon. Sam will start assembling a second OMC sometime soon as well.

October 5, 2007

• Tobin fixed the PSL intensity servo cables, including those going into the DAQ, and got the servo working again after Rana's mods. He took a transfer function from PD sensor to CS drive, and now he's working on a calibrated RIN measurement. He also re-enabled the ISS loopswitch in the "mcup" script.
• Pinkesh installed a new compute server named Mafalda.ligo.caltech.edu, and set it up with matlab and various ligo tools.
• Justin has upgraded the installation of mDV at the 40m to the latest version.
• Justin is trying to measure the mode cleaner transmission. He measured the power of the beam into the vacuum at the PSL table, and the power at the MC transmission pickoff port. But we don't know the reflectance of the MC transmission pick off. We're trying to figure it out from what we know about the pickoff mirror in the vacuum chamber.
• Rob has been working on locking the full dual-recycled FP Michelson configuration. The DRMI locks easily, and the hand-off to double-demod RF signals works fine, as do the fast triggers. Adding the arms has been causing lock loss. Tobin fixed the ISS, and that seemed to help. Rob is debugging locking scripts.
• Rana is working on a Bench model for a reconfiguration of the 40m for prototyping the new AdLIGO control scheme. He is considering the use of lighter ITMs to make the optical spring frequency higher and potentially visible in the noise curve. He's also considering whether we will need fixed or suspended folding mirrors in a long PRC.
• Bob and Luke had some challenges while assembling the AdLIGO Faraday isolator from Florida, which has some very strong magnets. They are improving the assembly and safety procedures.

September 28, 2007

• Elena presented the results of her simulation of cancelling seismic noise in the PRM configuration in the 1-10 Hz band, through adaptive FIR filtering of data from two 3-axis geophones. The adaption is via the LMA algorithm and the data from the six geophone channels are combined in a MISO configuration. The results are very promising! Elena is writing up her final paper.
• Tobin is working on getting the PSL intensity stabilization servo working again, after Rana's mods.
• Justin Garifoldi is visiting from LHO. He inserted a microscope slide in the PSL beam just before the periscope, to measure the beams going into and coming out of the vacuum system so that we can determine the Faraday Isolator isolation ratio. Right now the in going beam sample is 142 mW and the returning beam is 25 uW.
• Rob did some work with the output matrix for the short degrees of freedom (MICH, PRC, SRC) to minimize the loop couplings.
• Tobin set up a delay line shifter for the 166 MHz RF sidebands, and set it to optimize the POX166 signal for XARM.
• The front-end controller for ETMX, c1iscex, has been having sync problems, requiring reboots more and more frequently (like, daily). Rana has reduced the number of channels it logs to data acquisition system in an effort to reduce the cpu load, but that hasn't helped.
• Pinkesh installed a new cpu server into rack 1Y3 for analysis of 40m data (sofia.ligo.caltech.edu), including processing of digital video frames of beam spots from our new gig-e camera.

September 21, 2007

• Rob has been locking the dual-recycled michelson (without arm cavities), using double demod signals (with SRC->PODD, MICH->APDD, PRC->SP133). The handoffs from the single-demod signals (PRC->REFL33I, MICH->REFL33Q, SRC->REFL166I) used for lock acquisition appear to work well. He's working on better defining the output matrix and diagonalizing the loops.
• In the process, Rob found that the front end ETMX controller was running late. He and Tobin will work on diagnosing and fixing the problem, maybe with the help of Alex. Rana suspects there is too much load on the front end processors due to the large number of fast channels being acquired.
• Rana implemented new suspension damping filters on the MC optics, with resonant gain only at the pendulum frequencies. There is a significant reduction in rms noise. We'll look into implementing similar filters for the POS degree of freedom of all optics. Sonia's damping gain optimization needs to be redone with these new filters; Andrey is working on it.
• Andrey is using the XARM_CTRL signal to optimize the damping gain of ETMX and ETMY, using his modifications of Sonia's scripts. He can't do the YARM until we get YARM_CTRL into the DAQ.
• Elena has come up with some SISO adaptive filters to cancel the newtonian noise from the suspension signal using the signals from the accelerometers. She played with different filter orders and found that filters of order 64 or 128 seem to work well. Now she's working on combining data from multiple accelerometers in a MISO scheme.
• Sam has finished assembly of the first eLIGO OMC, and measured some of its optical properties. He finds a 96% mode matching coupling, 95% transmission for the carrier, and a finesse of 400. Compared to specs, this suggests an anomalously high mirror loss, despite drag wiping the mirrors and running the HEPA filter at full blast.
• Bob and Steve replaced the flaky 4-pin lemo cables on the ETMX and ETMY suspension controllers that go from the OSEM whitening boards to the AA chassis.
• Rana has received a gig-e camera for profiling beams, and Pinkesh will look into implementing it. The ultimate goal is to replace all the analog cameras at the sides and the 40m with a network of these things.
• Rob's talk on DC readout at Amaldi was chosen as a highlight paper for CQG. So now Rob is writing the paper.

September 14, 2007

• Rob is returning to the task of locking the full dual-recycled FP Michelson configuration in order to measure laser and RF oscillator phase- and intensity-noise couplings in both RF and DC readout.
• Tobin, Rana and Andrey were trying to lock the DRMI and they needed more RF sideband power. Using the optical spectrum analyzer, they adjusted the modulation depth for both the 33 and 166 MHz sidebands to get Gamma = 0.22.
• Tobin found and fixed a broken cable that prevented us from reading the N2 pressure. Now it reads correctly, and under-pressure will trigger a vacuum interlock. Steve tested this and it worked perfectly. IN the process, we rebooted the c1vac1 processor for the first time since Feb 2006. Rebooting has no ill effect on the vacuum controls.
• More computer problems: The iscepics process on the c1dcuepics computer died spontaneously. Tobin restarted it, but if it keeps on failing, we may need to get another computer to run it on.
• Andrey re-worked the video switch MEDM screen so that it is more intuitively oriented.
• Elena continues to develop her adaptive filter for feed-forward noise cancellation of ground-motion.
• Valera is here, to work with Sam on the eLIGO OMC construction. Sam is gluing up the optics and running tests.
• Steve replaced the tp3 dry scroll pump, after over 8000 hours of service.

September 7, 2007

• We had two (unrelated) computer failures this week, both of which brought down the suspension controllers. Alex traced the first, on Friday, to a power supply failure on the c1dcuepics computer; he replaced the supply and the systems came back to life. The second was on Tuesday, and Alex traced it to a failed disk on c1iscepics. He pulled out that computer, transferred the iscepics process from it to c1dcuepics, and brought it back up. Rana did a BURT restore to get everything working again. Alan got the nightly disk backup working; the second time, we needed help from Dan Kozak.
• Rana is exploring the origin of the RFAM that we now know is present after the Mach Zehnder (and presumably caused by it). We now have a reasonably reliable monitor of it in the MC reflected beamline.
• Elena and Andrey are working on understanding the coherence between accelerometers on the floor and the suspended optic motion, in order to implement some feed-forward noise cancelling.
• Steve is diagnosing problems with the vacuum N2 readback, which should trigger an interlock if it falls too low, but doesn't.
• SURF Elena Gaspari and grad student Andrey Rodionov went through the 40m safety training with Steve.
• Valera is coming on Monday to work on the autoalignment system and the OMC.
• Bram is coming to work on the tip-tilt mirrors for eLIGO.
• Rob and Tobin are back from Burning Man with lots of stories.

August 31, 2007

• Edward is finishing up his work on learning to use the new FTIR machine and develop a library of spectra for common contaminants, and is writing up his final paper. He and Bob are preparing to cross-calibrate with outgassing RGA spectra.
• Sonia is finishing up her work on optimizing our suspended optics controllers, and is writing up her final paper. All ten suspended optics have diagonalized input matrices. Most show great improvement over the previous diagonalization settings, although ETMX and MC2 not as good as the others. She has completed the optimization of the damping gains on MC1, MC2, MC3, by minimizing the rms of MCL.
• Keita, along with Rana and Rupal, have been searching for the origin of RFAM at the double-demod frequencies of 132 and 199 MHz, which are causing large offsets at the double-demod photodiodes, preventing them from being used to lock the dual-recycled interferometer. Keita observed the RFAM at MC transmitted port using a broadband-modified RFPD, and initially saw no evidence for it on the beam input to the MC.
• This led to much speculation about how the MC could generate such large RFAM. Keita and Andrey measured the length of the MC using the 166 MHz sidebands, and found it to be 4 kHz off, or 300 um too long. They adjusted the RF frequencies to get it centered on the MC length. It didn't help. Nothing else Keita tried, including rotating the polarization, seemed to help reduce the RFAM.
• Finally, Keita checked the RFAM at MC reflected port with the MC unlocked, and found that the RFAM was on the input light as well. So the problem isn't the MC, and Keita is going back to the usual suspect: the Mach Zehnder.
• Rana did some housecleaning, clearing out old PEM equipment (magnetometers, seismometers) that we weren't using (or even aware of).
• Steve did some housecleaning, clearing out old computers and monitors.
• Rob and Tobin are at Burning Man.

August 24, 2007

• The 40m SURF students all gave great talks last week, and Keenan, Alice and Darcy have moved on. Sonia, Josh and Edward are around for another week or two. We'll post the talks and final papers from everyone, soon.
• Sonia has measured and updated the input matrices for the OSEM sensors on most of the suspended optics controllers.
• Sonia has completed her optimization of the OSEM damping gain of the three MC optics, by minimizing the rms signal from MC_L in the 1Hz (pendulum) and 3Hz (stack) frequency ranges. She's making final measurements and tests, and writing up her work.
• Rana and Keita, debugging the DRMI lock instability, traced it to glitching in the BS UL OSEM, and determined that there is scattered YAG light on the osem. They are searching for the source of this scattered light.
• Tobin and Rana looked at the double-demod RFPD signals and found large offsets, presumably due to RFAM from the PSL.
• Keita and Rana have been working on finding and reducing this RFAM. In the process, they realizned and reoptimized the PMC, but it didn't help with the RFAM. They are now looking more closely at the Mach Zehnder, the possibility that the beam is not centered on the photodiodes, and the alignment into the mode cleaner.
• Rana execized the DRMI locking procedure; it still works, but there is lots of noise at the pendulum resonances, so we need to use Sonia's damping gain optimization on the BS/PRM/SRM.
• Tobin and Josh locked the XARM, YARM, PRM and then PRFPMI, centering all the relevant oplev beams onto their QPDs. They checked for and found no clipping of the oplev beams.
• Rob and Josh added some new fast OMC DAQ channels to monitor the OMC PZT. Rana added some 256 Hz SUS channels. Rolf is putting together a GUI for managing the DAQ channel configurations.
• Keenan moved the accelerometers to the BS support structure to see if they can be used for seismic noise cancellation. He and Elena took spectra and coherences.
• Rana is working on the ISS to upgrade its DAQ outputs.
• Rana used the 'SLOWscan' script to check the NPRO for multimode behavior. He left the SLOWDC slider in a nice place, far from "mode-hopping land".
• Tobin continues the debugging of the auto-dither-alignment system for the main interferometer (c1ass). He did two low-frequency swept-sine measurements to look at the transfer function between excitation of ITMX and ETMX and the c1ass error signals.
• Andrey Rodionov is a Caltech grad student getting his feet wet at the 40m. He has installed a new microphone to measure acoustic vibrations near the AS sensing table, and configured the ADC to get a good signal into the DAQ system.
• Steve notes that the Dycor RGA on the main 40m vacuum is rapidly dying. He is researching the purchase of a new one.
• The pressure transducer on the vacuum valve N2 line doesn't work, so when the N2 ran out it didn't trigger the software interlock which should turn off the main valve (V1). Steve is working on replacing it. We're also checking the correct behavior of the software interlocks under various pressure or failure conditions.
• Bob is back from LLO, having delivered and installed one of his bake ovens there.
• Stefan and Seiji are writing up their studies of various AdvLIGO and 40m configurations using Optickle. Seiji has returned to Japan.

August 9, 2007

• The 40m SURF students went off to LLO, and seemed to have had a grand time. They all continue to prepare for their final papers and talks, next Tuesday. Sonia has installed new suspension input matrices for MC1, MC2, MC3; they are much more diagonal. She is now optimizing the damping gains by minimizing the rms signal from MC_F.
• Tobin is back to the alignment servo comissioning (c1ass). He's taking swept-sine transfer functions.
• Sam is preparing to dirty-assemble the eLIGO OMC in Bob's clean room.
• Bob is off to deliver a bake oven to LLO, back in 2 weeks.
• Stefan and Seiji are developing an Optickle model of the 40m, including the application of RF sidebands with a Mach Zehnder, to plan for a test of the new AdvLIGO sensing and control scheme.

August 2, 2007

• We decided not to vent for drag-wiping, out of fear that it will reduce the arm cavity losses to the point where they are almost critically coupled, which will make locking much harder. We'll wait till Rob finishes his DC readout transfer function measurements.
• Rob has returned to locking the full IFO in power-recycled FPMI mode. He had to tweak the alignment of the MC optics, and fix a few other problems, but it all seems to still work fine.
• All the 40m SURF students have prepared abstracts and progress reports. Next week, they're off to LLO. Sonia is testing a new set of OSEM input matrices for the three mode cleaner optics. Next: optimizing the gain, automating the procedure, and moving on to the core optics.
• Alice has completed her noise analysis of the new RF photodiode design, and is preparing to assemble a prototype for testing.
• Josh continues to develop his code for automating recognition of higher order transverse modes in the PMC (and later, OMC) transmission scans.
• Keenan gave us a nice talk on the status of his gravity gradient analysis; see his slides.
• Steve and Josh continue to refine their photographic skills in capturing resonant spots on the test masses in exquisite detail.
• Steve removed a faraday isolator from Go's squeezer and returned it to Nergis.
• Bob is expecting a new shipment of powerful magnets to build a new Faraday isolator for eLIGO.

July 26, 2007

• We still plan to vent next week (Monday or Tuesday) to drag-wipe ITMX and ITMY. Josh has been taking lots of photos using the CCD cameras with the laser resonant in the arms, and also in air with the Nikon D40.
• We may also install the new DC photodiode preamp during the vent.
• Sonia continuies to take data with undamped suspended optics to diagonalize the input matrix, a necessary step before optimizing the damping gains.
• Alice has been taking transfer functions of the WFS quandrant photodiode.
• Our PSL head temperature started getting unstable earlier in the week, peaking as high as 45C and tripping an alarm, but in the last 5 days it seems to have gone back to normal, stable at 18C. Steve is watching it closely.
• Our frame builder computer (fb40m) has been running happily since Alex replaced the system disk last week.

July 19, 2007

• The system disk on our frame builder computer (fb40m) is dying, and fb40m has been crashing repeatedly for the last week; we have almost no frames during this period. Repairing it has not been easy; Alex finally diagnosed the disk as bad, but it took days to get a replacement (plus a spare, to set up a RAID1 system for this critical machine), and then had trouble copying the old, failing disk to the new one(s). Finally, Alex was able to replace the hard disk, and the frame builder is up and running again. Thanks, Alex!!
• We plan to vent next week (Monday or Tuesday) to drag-wipe ITMX and ITMY. Steve and Bob have prepared teflon "skis" to aid in carefully moving ITMX so that Tobin can get to the coated surface to drag wipe it. Josh will take photos before and after, using the CCD cameras with the laser resonant in the arms, and also in air with the Nikon D40.
• We may also install the new DC photodiode preamp during the vent.
• Bryan (from Australia) has analyzed the XARM ringdown data he took just before he left (before and after the drag-wiping of the ETMs). He also measured the mode cleaner ringdown time constant in order to properly deconvolute it from the coupled response. With these data, he was able to extract the finesse of the ZARM before and after wiping (1177 -> 1187). Analysis of the statistical significance of these results is ongoing.
• Rolf is preparing a proposal for installing a new CPU architecture at the sites for eLIGO, to be prototyped over the next year at the 40m. This new architecture should be much faster and cheaper than the VME cpus we now use, and should have much better capability for supporting testpoints (currently a big problem for us).
• With fb40m back up, Sonia is taking data for optimization of the suspension controllers.
• Bob and Edward have been trained in the use of the new FTIR system (for analyzing residual contamination of large in-vac parts), and they are working on commissioning the system. Edward gave us a nice talk on the system.
• Jamie brought over a driver he wrote to get data out of a GPIB device via a GPIB-to-USB interface board (which we bought a few weeks ago). Alice made it work with our ancient (but fast) HP4195A spectrum analyzer, and is preparing to use it to test RF photodiode response.
• Bob is preparing to ship a 24inch vacuum bake oven to LLO.

July 12, 2007

• We gave tours of the lab to 45 HS students from the Summer Science Program in Ojai (last week), and a dozen exchange students from Zambia (this week).
• Last week, Rob took spectra of the RF readout frequency noise coupling, before and after the drag wiping of the ETMs. As expected, the frequency noise coupling has decreased significantly.
• This week, Rob is at Amaldi, giving talks on DC readout at the 40m and also on the stochastic radiometer search.
• We are planning to vent once again, next week (Wednesday), to drag-wipe ITMX. We may also try to drag-wipe ITMY, but it is difficult to reach. Rana has a plan for moving ITMY within reach using teflon skis.
• We have a new and improved DCPD in-vac preamp, tested by Ben and soon-to-be tested again by Tobin. We could install it during next week's vent, but we'll probably wait till Rob gets back from Amaldi.
• Our framebuilder, fb40m, is dying; when it crashes, we lose data. Alex is working on diagnosing and fixing the problem. Sonia needs the DAQ to take data for her suspension optimization work.
• Sonia found that she couldn't optimize the damping on the suspended optics without first diagonalizing the OSEM input matrix, so she's working on that. She's writing a matlab script to fully automate the process.
• Alice has completed a noise analysis of her new RFPD preamp design, and finds improved SNR with respect to previous designs. She is getting ready to build a prototype circuit and test it with Jenne's AM laser.
• Josh has been frame-grabbing images of the test mass optics illuminated by the laser, and is analyzing them.
• Josh is developing an automated mode scanning system, and is testing it out on the PMC. Next, the OMC.
• Keenan is continuing his analysis of gravity gradient contributions to the LIGO noise. The effect of BSC chamber shaking is negligible. Next is the concrete slab.
• Jamie Rollins is visiting, and is currently working on photodiode characterization.

June 29, 2007

• Surf students Sonia Buckly, Alice Wang, Keenan Pepper and Josh Weiner were fully trained for all safety aspects of the 40m lab.
• All the SURF students are busily at work on their tasks, and have written first progress reports.
• Surf student Edward Macauley has arrived to work with Bob on the new FTIR machine. He'll get all the usual lab and laser safety trainings.
• After checking the alignment and noting all the reference oplevs and initial pointing, we turned off the in-vac PZTs and vented the 40m on Monday. A crew led by Tobin (including Darcy, Josh, Nick Smith, Rana, and Steve) took some closeup pics on the ETMs, applied light blasts of compressed nitrogen from a deionizer, carefully drag-wiped them (removing all of the dust particles visible by eye), and took more picks. They sealed up on Wednesday and pumped down (slower than usual to avoid steering up dust) by the end of the day.
• Tobin then repeated last week's pre-vent arm cavity loss measurements. It appears that the y arm is unchanged but the x arm shows a significant improvement. x arm: 216 ppm -> 152 ppm (significant improvement); y arm: 176 ppm -> 173 ppm (no change).
• Rana says: It's strange that there was an asymmetric change but dust knows no logic. One possibility is that the ETMX happened to have a large particle right in the middle of the optic where the beam goes. Very roughly we might expect that to lose in power we would need a particle with diameter ~ 0.05-0.1 mm, of which there were many.
• Josh took some pictures of the ETM's the Friday before the drag-wiping using the frame grabber on the CCD video feeds, and then took after-photos on Wednesday of this week (6/27). The changes in the pictures seem to agree with Tobin's before-and-after measurements. The ETMX camera showed a noticeable reduction in the scattering off of ETMX whereas the ETMY images were not as different when inspected visually.
• Ben reports that the new in-vac electronics for the DCPD has been delivered.

June 22, 2007

• Bryan is measuring arm cavity ringdowns in order to characterize the cavity loss, prior to venting.
• All the SURF students got laser safety training and 40m lab safety training.
• Keenan is working on modeling gravity gradients due to chamber vibrations.
• Sonia is measuring the damping time of our suspended optics, with the goal of optimizing the damping gain.
• Josh is working on implementing a frame grabber to record transmitted higher order modes of the output mode cleaner, and is learning how to model them.
• Sonia is learning about the RF photodiode preamp electronics.
• Darcy is working with Tobin on controlling the TCS system for eLIGO.
• Nick is visiting from MIT, working with Sam on the eLIGO OMC.
• Rob is locking the ifo in the PRFPMI configuration to work on reducing loop noise.
• Rana and Tobin tweaked up the alignment into the PMC and the reference cavity in the PSL, improving the transmission significantly. They plan to tweak up the FSS loop, etc.
• We plan to vent on Monday, to drag wipe the ETMs.

June 15, 2007

• SURF students have arrived and are coming up to speed; please make them welcome!
  Sonia Buckley of U Dublin will work on optimizing the suspension controllers.
  Josh Weiner of Caltech will work on analyzing the modal content of the OMC transmitted light.
  Alice Wang of Georgia Tech will work on aLIGO RFPD development.
  Keenan Pepper of U Florida will work on gravity gradients.
  Edward Macauley of Imperial College (arriving in 2 weeks) will work on FTIR testing.
• Bryan Barr is visiting from Glasgow.
• IFO commissioning, Electronics, controls, computers:
  • Tobin installed a delay line phase shifter in the mode cleaner rack, to facilitate the measurement of RF phase noise couplings.
  • Rana installed Shourov's code to make Qscan spectrograms of 40m channels.
  • We are preparing to vent in order to drag-wipe the test mass optics.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
  • Osamu and Go measured the suspension actuation function up to 10 kHz. They realized that there's an extra pole at ~ 5 kHz due to an RL in the coil / driver. Rana points out that there are other features in the 10 kHz region. Anyway, Go used the extra pole to correct the calibration of the shot-noise-limited sensitivity of the SRMI, for the squeezing paper. He will assign a ~ 10% error on the calibration, which doesn't impact the main point of the paper: that the shot-noise-limited sensitivity is reduced when squeezed vacuum is injected.
• Lab Infrastructure, Bake lab:

June 8, 2007

• There was a meeting of the 40meter Technical Advisory Committee (TAC) on June 7th at 08:30 Pacific. Rana went through these slides. The focus was on DC readout, and plans for the future.
• IFO commissioning, Electronics, controls, computers:
  • Rob got the interferometer back into reliable PRFPMI locking, after fixing some problems that was hampering lock acquisition recently. He noticed lots of glitches in the length sensing signals, and traced it to the c1ass alignment system.
  • We are preparing to vent in order to drag-wipe the test mass optics, in hopes of learning about anomalous losses and point scattering centers on the optic surface. Helena is giving us a tutorial on safe drag wiping. Steve is checking and turning up the HEPA filters.
  • The laser power has been slowly decreasing over the last couple of months; Steve is watching it carefully.
  • Rob noticed that the PSL autolocker is getting unreliable, and Rana is tracing it to some problems with the alignment into the PC and PMC cavities. To be tweaked up.
  • Tobin compared the RIN of our various HeNe lasers with that of the diode lasers we used to use for oplevs. The diode lasers are much worse.
  • Various scripts are being moved off of op140m onto other controls computers, with the goal of decommissioning op140m soon. Several EPICS processors had to be reconfigured to point to linux1 instead of op140m.
  • op540m continues to spontaneously crash, and suspicion lies in the dual-head video card. Under investigation.
• IFO modeling:
  • Stefan is modeling the new length sensing scheme for aLIGO, and planning is continuing for prototyping the new scheme at the 40m.
• DC detection:
  • Rob and Rana made a suite of measurements of laser and oscillator noise couplings in DC and RF readout, in the PRFPMI configuration. Some effort was necessary to compensate for the RFAM stabilization circuitry. Most of the broad features are understood, but some discrepancies in comparison with the Optickle modeling point to couplings that aren't yet understood.
• Vacuum squeezing:
• Lab Infrastructure, Bake lab:

June 1, 2007

• There will be a meeting of the 40meter Technical Advisory Committee (TAC) on June 7th at 08:30 Pacific. At Caltech, we will meet in the SCR. All interested parties are invited. TAC chair Ken Strain says: "Due to significant changes that have happened or are happening soon, we should try to take a long term view of the 40m plans this time round. In particular we should look at how the tests needed for the current projects (such as DR readout, new sensing schemes) fit with longer term activities (squeezing, other configurations work). Of course we should also have the technical updates as usual."
• IFO commissioning, Electronics, controls, computers:
  • Tobin and Rana worked on restoring lock of the power-recycled FPMI with DC readout. They ran into a variety of snags, mostly associated with alignment (and some electronics noise from POX), so they realigned the optics and the oplevs, and are almost back to full lock; work in progress.
  • Tobin built a little contraption to measure RF oscillator amplitude noise, and took a spectrum of the relative amplitude noise of the Marconi RF generator from 100 Hz to 100 kHz; it falls from 5e-7 to 5e-8.
  • Tobin and Steve did a shootout of three different HeNe lasers, measuring their RIN for use as oplevs. The new JDSU model 1103P had the lowest RIN by far (10^-7) up to 550 Hz. Sam suggested repeating the test while swapping power supplies to see if that's responsible for the large differences in RIN between the different lasers.
  • Rana is working on retiring op140m by moving all the directories off of it; Alex will move the YP server off of it, and then it will be shut down. If it isn't missed, the box will be removed and used as spares for the computing group.
  • Bob tested the eLIGO OMC glue joints made by Sam; they all look good on the RGA, and they are now in the cavity ringdown test system. Sam will be working on the SP table to set up some OMC optics tests.
  • Rob returned from S5 shifts to find that the RGA logger was down. He restarted it, and he suspects a flaky cable or connector on the iServer, or on the ethernet hub in the EY rack.
• IFO modeling:
• DC detection:
  • Rob has been on S5 shift at LLO, so no progress here.
• Vacuum squeezing:
  • Go is writing his thesis and preparing final draft of the squeezing-enhanced interferometer paper.
• Lab Infrastructure, Bake lab:
  • We are preparing for the arrival of 5 SURF students, and making sure thare are working PCs for them.

May 25, 2007

• IFO commissioning, Electronics, controls, computers:
  • After last week's dewhitening board modifications, Rob measured the noise at the coil drivers for the face coils of our four test masses at around 300Hz (near the minimum of the new dewhite transfer function), with SUSPOS damping on. As expected, there is significantly less noise with dewhitening on.
  • Tobin encoutered test point problems with the dither alignment system (c1ass). Alex's expertise is needed.
  • Steve continues to improve his photography skills, capturing clear pics of the pattern of point scatters on all four test masses with the arms locked. Cheryl Vorvick at LHO has been helping to interpret the pics, identifying the same point scatters from two different views of ETMX.
  • After Steve completes his photographic odyssey of the point scatters on the test mass optics, we plan to vent in the next couple of weeks to drag wipe the optics.
  • Steve replaced the oplev HeNe laser used for BS/PRM/SRM with a new JDS one. He is now retuning the spot diameters on the oplev QPDs.
  • We ordered a new PC for the SURF students (Dell Optiplex).
  • We would like to retire op140m, the Sun E450 "refrigerator". We'll need help from Alex to decommission the services it performs.
• IFO modeling:
• DC detection:
  • After modifying the test mass dewhitening filters, Rob took some DC readout noise spectra with the PRFPMI. There is significant noise reduction in the 80 - 400 Hz region. He implemented a "cheesy" MICH feedforward path (feeding the MICH signal to the ITMs), and made further improvement in the noise between 80 and 300 Hz. The spectrum above 800 Hz is unchanged, bottoming out at 2e-18 m/rtHz at 1 kHz.
  • Rob has been taking a series of measurements of transfer functions from noise sources to DARM with DC readout and RF readout, including laser intensity noise and RF oscillator phase noise. Interpretation of the plots are in progress.
  • The DCPD is stuffed, and in its pipe nipple. Ben is almost finished testing the DCPD, and should be delivering it early next week. Steve will order a new pair of end flanges for the nipple, and Bob will get the whole lot cleaned and baked.
• Vacuum squeezing:
  • Go (with help from Osamu via videophone) retook a noise spectrum with the SRMI and redid the calibration. The noise is 8e-17 m/rtHz at the shot-noise limited frequency range around 10 kHz. This was the remaining measurement before finalizing the squeezing-enhanced paper and releasing it.
• Lab Infrastructure, Bake lab:
  • Rob and Steve gave a tour of the 40m to a bunch of Chilean student visitors.
  • Bob has completed the vacuum prep for the LASTI SEI system, and has now moved on to the SUS noise prototype for LASTI, and the housing for the eLIGO DC PD amplifier.

May 18, 2007

• IFO commissioning, Electronics, controls, computers:
  • Rob has finished modifications to the dewhitening boards for the test mass coil drivers, including lower noise opamp and a new PZ network for more dewhitening. Tobin measured transfer functions and offsets with the dewhitening switched on and off, and verified correct functionality.
  • After re-installation of the modified dewhite boards, and a bit of twiddling, Rob got the IFO locking again in low-noise PRFPMI mode. It handled the dewhite-switching without a hitch.
  • Steve finished installation of a new HeNe laser for the BS/PRM/SRM oplevs, and is doing final adjustment of the spot sizes on the QPDs.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
• Lab Infrastructure, Bake lab:
  • Sam has been doing some glass-on-glass epoxy tests for the ELI OMC assembly. He gave various glued optics to Bob for vacuum RGA tests.
  • Bob is nearing the end of vacuum prep of SEI parts for LASTI.

May 11, 2007

• IFO commissioning, Electronics, controls, computers:
  • Rob made some laser intensity noise measurements using a single bounce off the PRM and the REFL33 diode. He finds a forest of peaks in the 1-2 kHz region that closely parallels our DARM noise spectrum. Rob suspects some clipping in the REFL path.
  • Rana made some measurements of the ITM and ETM OSEM coil driver contribution to our DARM noise in the 100 Hz region. He and Rana have pulled the dewhite boards and are putting in new lower noise op amps and newly designed PZ dewhite filter stages.
  • Steve has been using our Nikon with a new zoom lens to take close-in pictures of the laser-illuminated test masses, at 0 and ~45degree view angles, in order to undersand the observed scattering. He has identified what appears to be a izeable (~ mm?) scratch in the middle of ETMX, visible only at ~ 45degree viewing angle.
  • Controls computer op540m is chronically misbehaving, presumably associated with the fact that it has two video cards and heads. We will need some help to make it useable.
• IFO modeling:
  • Rana and others have been thinking about modifications to the 40m to test the new AdLIGO RF sensing scheme, featuring a much longer, folder PRC, and much lighter ITMs to enhance radiation pressure effects which will be important with the higher laser power in AdLIGO.
  • Rana did some modelling of the effect of the Krypton from our leaky DCPD nipple on the noise. It is surprisingly large, and he says it would be neat to leak some more in our vacuum to see the noise in DARM.
• DC detection:
  • Rob measured the relative coupling of laser frequency noise to the DARM signal, in RF and DC readout, in a PRFPMI. There is a 20 db reduction in the noise coupling with DC readout above ~ 600 Hz.
  • The new DCPD board is stuffed, and in its pipe nipple. Ben started a functional test last week, and he'll finish it up next week, when he gets back.
• Vacuum squeezing:
  • Go and Nergis have a near-final draft of a paper on the 40m squeezing experiment, which is intended for Nature Letters.
• Lab Infrastructure, Bake lab:

May 4, 2007

• IFO commissioning, Electronics, controls, computers:
  • The PSL has been free of glitches (in temperature or power) for the last few weeks.
  • Steve put in some optics to reduce the PRM and SRM oplev beam spot size on their QPDs. New oplev lasers for the ITMs are on order, mounts and optics are in hand (mostly).
  • As an aid to understand the anomalous losses in the arm cavities (at the 40m, and by extension, the sites), we will vent to drag wipe the test mass optics in the next few weeks. Tobin and Rana took "before" measurements of the arm losses. Steve, Rana and Rob have been honing their photographic skills, trying different cameras and lenses to get good, hi-res pics of the beam on the test masses. Rana ordered a framegrabber so that we can easily make digital pics of the optics with our video cameras.
  • Tobin continues to commission the dither-alignment system. There is a new EPICS word derived from the lock acquisition state vector, to assist the system in determining which cavities are in lock and available for dither-alignment.
• IFO modeling:
• DC detection:
  • Rob is compiling a list of measurements to be made of noise couplings to DC and RF readout of DARM, from laser and oscillator amplitude and frequency noise, MICH and PRC loop noise, Mach-Zehnder noise, and input and output PZT steering mirrors. He's making cables and connectors to facilitate measurements of transfer functions.
  • Ben has a completed DC photodiode board. Rob will test it.
• Vacuum squeezing:
  • Go and Nergis have a first draft of a paper on the 40m squeezing experiment.
• Lab Infrastructure, Bake lab:
  • The office remodeling is complete. The flow of chi is optimized. Tabula rasa reigns.
  • Bob continues to bake parts for the LASTI SEI system; he sees the light at the end of the tunnel.

April 26, 2007

• IFO commissioning, Electronics, controls, computers:
  • Tobin and Valera are commissioning the dither alignment system: Automation scripts being written, code being modified to enable sensing of the IFO lock state. Works for the arms and PRM is being worked on.
  • BS / PR / SR optical lever diode lasers were replaced by a single HeNe and a few beamsplitters. HeNe telescope was optimized to maximize the angular sensitivity. There is a factor of about 10 suppression of the error signals at 1 Hz. Along the way we realized that the OL noise was dominated by wind and so we'll investigate acoustic damping of the tables as a summer student project.
  • After the heroic efforts of Steve and Valera to make our new PRM/SRM/BS HeNe oplev as good as the old diode lasers (and certainly soon to prove itself better), we can now lock the IFO in the daytime again.
  • We are procuring parts (lenses + 1 HeNe) to outfit the ITMs with HeNe oplev lasers.
  • Following Rana's suggestion, Tobin and Valera tried using the beam transmitted by the MC2 to damp the eigen mode of the three mirror cavity which is not being sensed by the MCWFS i.e. the one that moves the beam on the MC2 but does not change the cavity waist position or angle relative to the input beam. The signals from the MC2 transmitted quad detector are going to the MC2 oplev screen. We closed the loop by feeding back to the MC2 mirror. Tobin is working on evaluating the loop shape and the noise.
  • We are gearing up to vent and drag wipe the test mass optics to try to reduce losses and measure the effects in the noise couplings.
  • A campus-wide power outage Monday morning gave Rob and Tobin an opportunity to exercise our coldstart procedures. After a little bit of struggle getting the framebuilder back up, everything came back up. Need to get the framebuilder & RAID array on UPS!
  • Our "network appliance" PC died. On advice from Alex, Rob replaced it with a Linksys NAT router, and it seems to work well. The backup of frames and code to the CACR archive needed a restart.
• IFO modeling:
• DC detection:
  • Noisy OMC computer was moved away from the AS port, but there was no change in IFO noise. The noise from 100-1000 Hz has a 1/f^2 shape and so we are beginning to measure more actuator noise.
  • The near term plan is to continue measuring the laser noise couplings of the DC readout + PRFPMI system.
• Vacuum squeezing:
  • Go is writing a paper on the squeezing-enhanced 40m.
• Lab Infrastructure, Bake lab:
  • We optimized the Feng Shui of the office space today to optimize the flow of chi. It is hoped that this will help us to understand the DRSE locking signals.

April 19, 2007

• IFO commissioning, Electronics, controls, computers:
  • Tobin continues to commission our new digital dither alignment system, and Valera is visiting, to help. Finding and fixing bugs in TDS, with help from Matt Evans. He remeasured the xarm sensing matrix using the oplev readback as a measure of the excitation, and it looks sensible. They are making good progress in implementing the servos.
  • Steve has been carefully monitoring the temperature and power of our MOPA laser, because of the power and temperature glitches that have plagued us in recent weeks. He reports that the MOPA has been running stably this week, with no sign of glitching.
  • Rana found and fixed some bugs in the FSS slow servo code. It made him feel better about it.
  • Rana is ordering a 4 input USB frameGrabber so that we can save camera images of our test mass optics before and after drag-wiping, and also to quantitatively mode-analyze the beams transmitted from the PMC, MC, IFO, and OMC. This will be a good SURF project.
  • Rob has got our new laptops working with dataviewer, dtt, and epics/medm.
  • Rana is moving autoscripts over to our "headless" control computer, op340m, and it seems to be working stably.
  • There are still a variety of problems associated with our controls computers. For example, only one computer (op440m) can perform excitations via dtt and see the relevant testpoints. Rana is working on having one computer serve all the LIGO controls applications, so that all controls computers can be configured identically (so either they all will work or none will work). This is proving to be more complicated than we had hoped.
  • Steve is getting longer ethernet cables to allow the relocation of the c1omc computer further from the AP table.
• IFO modeling:
• DC detection:
  • Rob is working on measuring the laser frequency noise coupling to DC readout, and is making new cables to facilitate these measurements. He's also working on the modeling of these couplings for comparison with the measurements.
  • Rob is working on a document summarizing the expected advantages of DC readout: elimination of RF intensity noise, RF frequency noise, RF mode overlap imperfection (up to OMC finesse); reduction of carrier junk light (up to OMC finesse). The document will enumerate the relevant noise couplings that he is measuring in both RF and DC readout for his thesis (laser frequency and intensity noise, RF oscillator frequency and intensity noise, loop couplings, etc).
• Vacuum squeezing:
  • Go is working on his thesis, and papers.
• Lab Infrastructure, Bake lab:
  • A film crew filled up the 40m lab last Friday in order to film an interview of Kip Thorne for the PBS series "Closer to Truth". Steve shuttered the laser during their filming, and when they left, everything (PSL, IFO) came back up nicely.
  • Visit by 60 Hamilton Elementary School 6th graders on Friday cancelled because of rain.
  • Steve is rearranging desks and workbenches in the office area, to better suit our needs.

April 12, 2007

• IFO commissioning, Electronics, controls, computers:
  • Rob and Sam have been exercising the lock acquisition for the PRFPMI, which has been working well. They had some trouble with the Y-arm violin modes ringing up during lock acquisition, so they adjusted the notch filters on SUS-ETMY_LSC, which quieted them down nicely.
  • In the PRFPMI configuration, Rob notes that the noise spectra look the same as they did a few months ago. There is still a 1/f^2 noise from 100 - 1000 Hz, of unknown origin. It is not reduced when coil driver dewhitening is turned on, so that is sub-dominant. Right now, Rob is focusing on understanding and eliminating the forest of peaks in the 1-2 kHz region, which is suspeced to be due to the input PZT steering mirrors.
  • Our MOPA (main PSL laser) has recovered most of the power lost during recent "events", and is now hovering above 4 watts. Steve notes that it has small (~0.1 watt) "hiccups" of lost power, recovering in minutes or hours. This seems to be associated with the NPRO pump diode head temperature. We are watching this carefully.
  • Steve is acquiring ethernet cables of appropriate length to move the c1omc processor away from the AS ISC table, since it is generating considerable acoustic noise.
  • Rob is getting the controls software running on our new laptops (with Fedora 6). He's got DTT and dataviewer working, but is having trouble getting EPICS/medm working. In progress.
  • Ben has stuffed the board for the new DCPD, and has worked around the American Capacitor part problem. He just needs to put on a connector, and test it out. We plan to install it into a new vacuum nipple and then into the vacuum chamber in the next few weeks (when we vent to drag-wipe the TM optics).
  • Rob is setting up the new "headless" controls computer, op340m, to run our standard scripts continuously.
  • We continue to have reliability problems with the op540m control computer. Nothing that a weekly reboot can't fix.
  • There will be an AdLIGO ISC meeting at Caltech next week. We will make plans for a redesign of the 40m to test the new proposed RF scheme for AdLIGO.
  • We are planning out projects for SURF summer students.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
  • Go and Osamu have been tuning up the squeezer after an SHG heater wire failed, changing the alignment of the SHG. They have gotten back to ~ 3 dB of squeezing at 1 MHz, and are continuing their optimization of modematching, alignment, etc to get to their goal of > 6 dB.
• Lab Infrastructure, Bake lab:
  • We will have 60 sixth-graders from the Hamilton Elementary School visiting the 40m next week.
  • Steve is rearranging the office space to make room for another electronics bench and a whiteboard.
  • Bob is quite busy with large bake loads for LASTI.

April 5, 2007

• IFO commissioning, DC readout, Electronics, controls, computers:
  • Tobin continues to commission the new dither alignment system. He and Rana concluded that the two input PZT steering mirrors are so close together that they are nearly degenerate and only control one pair of angular degrees of freedom. They also determined that they need some kind of digital bandpass filter before the digital demodulators. Tobin will generate the appropriate code in Bork-space. Rana also suggested that the oplev readbacks be used to determine the sensing/control matrix.
  • Rana found that the SDSEN damping of our suspended optics wasn't working because of some bug in the gain ramping: with the TRAMP field set to a non-zero value, the ramp time for the SDSEN gain becomes infinity (or close to it), so no change to the gain will work. He set TRAMP to 0 for all SDSEN damping loops and now they all seem to be damping better. He suggests that we systematically optimize the gains of all suspension damping loops, using Justin's newly-installed matlab hooks. A good SURF project!
  • Rob is working on the dewhitening boards for all the test mass coil drivers. He's measuring the noise with dewhitening off & on and diagnosing bad opamps.
  • Our main laser (MOPA) continues its slow recovery from the dramatic power loss of a few weeks ago. The power is now back up to 4.2 watts.
  • Rana talked with Gregg Harry about optic charging and UV light discharging. They decided to hold off on any tests using 40m optics until the procedure is well tested elsewhere and deemed safe for the mirror coatings.
  • Rana talked with Bill Kells about the scatterometer tests Bill was proposing. They agreed that the arm loss measurements already done by Tobin, as well as hi-resolution pics of the optics are all we need before venting to drag-wipe the test masses and repeat the measurements. This will determine whether the anomalous cavity loss is due to dust or coating imperfections. If it is the latter, Bill will proceed with scattering measurements. We plan to vent for drag-wiping in 3-4 weeks, depending on when Rob is finished with his next round of DC readout measurements.
  • Tobin took a 9-sec exposure of the PMC with our new D40 digital SLR camera and determined that it sees 1064nm light as purple, with quite high resolution.
• IFO modeling:
• DC detection:
  • Rob is gearing up for a series of measurements of noise couplings, comparing DC readout to RF. First up is laser intensity and frequency noise, and RF oscillator intensity and phase noise. Then loop noise couplings, angular (oplev) couplings, etc. He's developing ways to drive the various noise sources in a way that doesn't saturate them and gets good coherence, is calibratable in physical units, and is do-able from the control room.
• Vacuum squeezing:
  • Go and Osamu have been tuning up the squeezer to achieve more squeezing at lower frequencies.
• Lab Infrastructure, Bake lab:
  • Bob got a newfax/scanner/color-printer to replace the one that hasn't worked in months. Christian will set it up and send instructions on how to use it.
  • Bob's bake lab is cooking on all burners, vacuum-prepping parts for LASTI.
  • Trees along the x-arm roof were trimmed.
  • A power transformer failed on Wednesday early morning, shutting down the AC. It will take 10-14 days to replace, so chilled water supply will be crippled. Expect the ifo lab to be warmer than usual.
  • A noisy AC unit on the south annex roof has been fixed.

March 29, 2007

• IFO commissioning, DC readout, Electronics, controls, computers:
  • Tobin continues to make progress in commissioning the new dither alignment system. He measured the control matrices, inverted them, and set up the servo matrix. The servo restores the proper alignment on the x arm pitch. Yaw, and y arm, come next.
  • Rana tuned up the MC WFS system.
  • Rob installed two stages of dewhitening, with "design" digital compensation, for all four TMs.
  • Steve installed an old HeNe laser for the BS,PRM,SRM oplevs. He and Rana will rebalance the power to each mirror, and will place lenses to get small spots on the oplev QPDs. Steve has ordered two new JDS/uniphase HeNe's for the ITM oplevs.
  • Following discussions with Garilynn and Bill Kells, we are planning to perform a series of measurements to determine the amount of scattering and cavity loss in the 40m arms, before and after drag wiping, which we plan to do in about 4 weeks. (cf T070051).
  • Rob and Tobin are looking at frame grabbers to document our mirror scattering before and after drag-wiping, and also to analyze HOMs in the OMC.
  • We have a new NPRO sent to us from LHO, as a spare. The NPRO in our main laser lost 30% of its power a few weeks ago, but seems to be slowly recovering. Rana and Steve upped the diode current, and we are now almost back to the power we had before.
  • Sam and the ELI OMC group are making plans to test OMC assembly, including vacuum compatibility of various epoxies, in Bob's bake lab. They expect to recieve the OMC fSi breadboard in 8-10 weeks. They will assemble the OMC and suspension in E Bridge 56.
• IFO modeling:
  • We're thinking about how to implement the new proposed AdLIGO modulation scheme (9/45 MHz) in the 40m.
• DC detection:
  • Ben has the new DCPD PCB, and will stuff it soon.
• Vacuum squeezing:
  • Go is repeating some measurements on the squeezing obtained with the PPKTP OPO, and responding to reviewers comments on the PPKTP paper. He measured the linewidth and FSR of the OPO cavity.
• Lab Infrastructure, Bake lab:
  • Steve checked and labelled circuit breakers & recepticles all over the lab. Daisy changed ac power connections were eliminated.
  • The Caltech 40m Bake Lab will be extremely busy for the next 90 days, cleaning and baking BSC ISI parts for LASTI. If you need to have something cleaned and baked please contact Dennis Coyne to have your job prioritized.

March 22, 2007

• Rob gave a nice status report on the 40m activities at the LSC meeting in Baton Rouge.
• Osamu gave a nice report on his e2e lock acquisition studies for AdLIGO arm cavities.
• Go gave a nice talk on the recent results from the vacuum squeezing injection experiment.
• Kirk gave a nice talk on plans for prototyping the new AdLIGO ISC scheme at the 40m.
• IFO commissioning, DC readout, Electronics, controls, computers:
  • Not much activity, since most people are at the LSC meeting.
  • Doug Cook sent us a new JDSU NPRO to replace our dying one. (JDSU model M126N-1064-700 SN 5519 from Hanford). It has been received and tagged. We are making plans for when to install it.
  • We had discussions with the AdLIGO optics group about performing some experiments to measure scattering loss with the 40m arm cavities, to determine whether the anomalous losses are due to coating inhomogeneities, dust, or something else. This may involve venting in order to drag-wipe the optics to remove dust. We are also looking into the possibility of illuminating our test masses with UV light to discharge them, potentially releasing the dust that sticks to their surface.
  • Steve installed an old Uniphase HeNe laser with two 50% beamspitters as oplevs for the BS, SRM and PRM.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
• Lab Infrastructure, Bake lab:

March 15, 2007

• IFO commissioning, DC readout, Electronics, controls, computers:
  • Last week, our NPRO power dropped by 30% (which we believe is due to the pump diode), and it hasn't recovered (but it hasn't gotten worse). The MOPA is now only putting out ~ 3 watts. Steve opened up the MOPA to visually inspect the NPRO, and didn't see anything suspicious. We need more power! We are looking for a spare NPRO.
  • Steve redistributd the remaining power so that a sufficient amount goes into the main interferometer, leaving the squeezer optics with almost nothing.
  • Ben built a latching relay box so that when the PSL enclosuire interlock is broken (by the enclosure door being opened without deliberately bypassing the interlock), it stays broken until it is manually reset. There's a problem with the bypass function, which Ben hopes to fix soon.
  • Tobin, Rob, and Sam exercised the "pringle" code that Alex installed here last week. They excited the pringle mode of the test masses and adjusted the gains to minimize the coupling to POS.
  • Sam and Tobin continued their commissioning of the new dither-alignment system. They turned up the dithers till they got coherence; adjusted the demodulation phases for the YAW half of the system, and automated the process. They measured the ITMX, ETMX, and PZT sensing matrices, and inverted to get the control matrix. However, the resulting system wasn't able to improve the alignment. More work is required.
  • Rob returned to PRFPMI locking, which went pretty smoothly. Started measuring some frequency noise TFs, but was stymied by testpoint/syncing problems in the c1omc.
  • Rob struggled with a slew of tespoint problems in the controls, necessitating some front-end cpu reboots. Not sure what is causing these instabilities. He also struggled with some flakiness in the communication between the EPICS slow controls and the front-end servos, necessitating a reboot of c1losepics. He will consult with Alex and Rolf.
  • Rob switched around some cables and RF photodiodes, so AS166 became AS33 and POX33 became POX166. He thinks this PD setup is more useful for us in the near term. We'll probably want an AS166 back when we go for RF-DC comparisons with the full DRFPMI.
  • Mike installed DST patches on our solaris machines, and Tobin took care of the linux machines.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
  • Go is in Japan. At NAOJ/TAMA, he gave a talk on progress with vacuum squeezing at the 40m. He'll give a version of this talk at the LSC meeting, and we plan on having a get-together there to talk about the recent results and the next steps.
  • As noted above, our dying laser leaves almost no beam for the squeezing setup.
• Lab Infrastructure, Bake lab:

March 8, 2007

• IFO commissioning, DC readout, Electronics, controls, computers:
  • Tobin has been commissioning our auto-dither-alignment system. He has written a variety of scripts to configure the controls, apply angular excitations (dithers in the 4-7 Hz band) to the suspended optics and PZT input steering mirrors, and use the system to digitally demodulate signals from the DC photodiodes. He tried some excitations and had trouble getting good coherence. Under investigation.
  • Our Main MOPA laser lost ~1.5 watts of power on Tuesday night, apparently due to a sharp drop in the NPRO power. This may be due to the NPRO pump diode dying, although so far, 2 days later, it's holding steady at the lower power. We have to be prepared with a source of a spare to replace our NPRO if it dies.
  • Tobin applied DST patches to the linux control computers in the lab, and Larry and Mike are doing the same for the Suns.
  • Rob and Alex are debugging a problem with autoburt causing it to die spontaneously.
  • Alex and Sam have installed code in our suspension controllers to excite the "pringle" mode of the optic (UL-UR+LR-LL), as part of a study of upconversion / Barkhausen noise.
• IFO modeling:
• DC detection:
  • Ben has re-designed the pcb layout for the DC photodiode preamp, incorporating some part choice suggestions from Rana. He'll send it out for fab by early next week.
• Vacuum squeezing:
  • Osamu and go were able to directly drive the beamsplitter at 50 kHz to simulate a GW signal in the signal-recycled Michelson (SRMI) configuration.
  • After a couple of nights of failing to lock the SRMI due to high microseismic noise, they were able to lock the SRMI, implement the noise-locking servo, and make measurements with injected squeezed vacuum.
  • They measured broadband squeezing-enhancement in the SRMI continuously from 40kHz to the end of the SR785 detection band (100kHz). They demonstrated an increase in S/N by decreasing the shot noise level by about 3dB and retaining the strength of a simulated GW signal at 50kHz.
  • The theoretically predicted shot noise level agrees with the measured noise level of the SRMI at frequencies above 40kHz. They also made sure that it's shot noise by doubling the interferometer LO power twice and showing it goes up by 3dB each time. At low frequencies below 30kHz, the noise level goes up by 6dB, which indicates it's laser noise or interferometer length noise.
  • They also calibrated the strain sensitivity of the SRMI using the known strength of the injected simulated GW signal.
  • They will be repeating their measurements and making more checks, but this looks like the real deal. The first squeezing-enhanced suspended-mass interferometer GW detector!
• Lab Infrastructure, Bake lab:
  • Steve completed an inventory and photo album of all the lasers in the laboratory.
  • Steve and Tobin installed several beam dumps on the PSL table, and checked the PSL enclosure interlock.
  • The PSL enclosure interlock does not latch when it is tripped off. Ben ordered a relay and switch for the PSL enclosure interlock so it operates in the desired manner. When these come in, he'll install them.

March 1, 2007

• Comings and goings:
  • Kirk is off to LLO next week, then back home to ANU (with a short stop in California). As a goodbye present, he supplied the group with some outstanding sugar napoleans during our group meeting. Thanks, Kirk! Come back soon!
• IFO commissioning, DC readout, Electronics, controls, computers:
  • Osamu noticed that XARM lock was unstable, and he narrowed it to the ETMX oplev servo. Rana noticed that the oplev laser (an old HeNe, installed last month) was dying. Steve replaced it with with spare, and is planning on buying new ones, from JDS. He'll start with one, and measure the intensity noise to see if it's suitable.
  • The ITMX side osem damping has never worked well, and is rather strongle coupled to POS. and Rana is complaining that it is hampering lock acquisition work. Osamu has been diagnosing it. He can get it to damp well by turning the gain up to 1e7. Something is probably wrong with the electronics / cabling. Work in progress.
  • Royal has completed her work on modifying all the oplev whitening boards on all 4 test mass suspensions. The boards were tested before and after installation, The noise spectra were dominated by ADC noise above 10 Hz; now 1 kHz.
  • Rana modified the ITMX and ITMY oplev filters. The loop gains are now higher; ~5 Hz instead of 2 Hz. And the filters give ~10 dB less noise from 20-100 Hz. Also there are new HVAC filters in both to account for the frequency shift in our AC motors.
  • Tobin calculated the round-trip cavity losses from the measurements he took last week: x arm: 228 ppm; y arm: 187 ppm. This isn't much worse than at the sites (~ 150 ppm rond-trip losses, according to Bill Kells), but it unacceptable for the high-finesse arms at the 40m (and AdLIGO). A great challenge for the AdLIGO core optics team!
  • Tobin has been thinking about how to commission our new auto-dither-alignment system.
  • Sam plans on exciting the "pringle" mode of the LLO test masses in order to study Barkhausen noise and upconversion. This requires modifications to the front-end SUS controller code. He wants to test these mods at the 40m.
  • Tobin has been working on AdLIGO TCS design and modeling with Phil.
  • Our new controls laptops now can see the wireless network, thanks to work by Rob. Still need to configure them correctly to run controls software.
• IFO modeling:
  • Kirk has been learning about, and modeling, the proposed new AdLIGO control scheme (9/45 MHz), and is thinking about ways of implementing the scheme at the 40m.
• DC detection:
  • The timing slave module for the OMC controls went bad. Jay replaced it with a new one, and the OMC controls are now happy.
  • The OMC controls were having a variety of problems associated with the test point manager, probably exacerbated by the timing slave problems. Rob is waiting for Alex to return, to help find and fix the root problem.
• Vacuum squeezing:
  • Osamu and Go installed a second Faraday rotator to suppress the optical feedback from the interferometer to the OPO cavity. It worked pretty well and they were able to lock the SRMI with DC-locking of the dark port field using the MICH offset.
  • Now with this stable DC power, we were able to measure reduced interferometer noise (at ~ MHz frequencies) with the injection of squeezed vacuum, with more accuracy and reliability. About 3dB of reduced noise due to squeezing was observed.
  • They can now noise-lock the squeezed vacuum phase more stably, to observe broadband noise reduction. They are working to establish the frequencies where interferometer noise is dominated by quantum shot noise.
  • They measured the noise floor of the signal recycled michelson (SRMI) at various input powers, and concluded that they were seeing quantum-limited shot noise above ~ 40 kHz. There is lots of excess broadband noise, and lines, at lower frequencies. They managed to shake the ETMs at 50 kHz (using analog inputs to the coil drivers) in order to crudely calibrate the spectrum.
• Lab Infrastructure, Bake lab:
  • Ken Barat, Advanced LIGO Laser Safety Consultant, got a tour of the lab and its safety procedures. We tested our PSL enclosure interlock: when the door opened, the laser was automatically shuttered, as desired. When the door was closed, the laser automatically un-shuttered - not as intended! We'll get the Abbotts to fix this.

February 22, 2007

• Mike Zucker livened up the lab for a couple of days this week.
• IFO commissioning, DC readout, Electronics, controls, computers:
  • Rana, Kirk and Rob continue to work on improving the locking of the PRFPMI, finding and fixing problems and noise sources.
  • Rob has been working on reducing the noise in the DC readout of the PRFPMI, by changing loop gains, etc. He worked on diagnosing the origin of the large noise peaks in the 1-2 kHz region of the DC readout noise spectrum taken last week. His theory that it was due to DAC noise on the PZT steering mirrors turned out to be wrong.
  • Royal installed the whitening board she built for the ITM optical lever signals, with improved filtering. Signals are now seen above the ADC electronic noise floor in the 10-1000 Hz frequency range. She also built and installed new cables.
  • Tobin, Kirk, and Rana changed the main drive resistors on all the TMs from 100 to 400 Ohms and saw the expected noise improvement of 4 at frequencies below ~ 300 Hz. (in the PRFPMI configuration with DC readout). They tried many things to reduce the noise in the 300-3000 Hz band, without success.
  • Tobin ran Rana's scripts to repeatedly measure the DC power with the arms in the locked and unlocked state, in order to get an estimate of the losses in the arms. Data are being analyzed.
  • There's something very wrong with the side damping on ITMX, which might be a mis-mapping of the OSEM cables. Osamu is investigating.
  • Steve will install a HeNe laser for the BS/PRM/SRM oplevs, in the next week. The HeNe lasers on the ETMs are much less noisy than the diode lasers we have been using. Steve will procure parts to replace the remaining diode lasers, on the ITMs.
  • There have been a fair number of problems associated with the controls computers, but Larry's group has been helpful in resolving them.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
  • Osamu and Go improved the alignment of the squeezed vacuum injection optics, and obtained much better mode matching into the interferometer. This in turn resulted in a much higher level of squeezing; they measured 3.5 dB of squeezing at 4 MHz.
  • Osamu and Go noise-locked the injected squeezed vacuum with the interferometer in the signal-recycled Michelson (SRMI) configuration. They observed a squeezing-enhanced noise spectrum at frequencies down to 4.5KHz, with about 2dB of broadband noise reduction. They excited the BS to simulate a GW signal at 7kHz and will use that to calibrate the spectrum. It is not certain that the noise is quantum-limited shot noise (for one thing, the spectrum isn't flat). Under investigation.
• Lab Infrastructure, Bake lab:
  • Rana ordered, received, and configured two new laptops for EPICS controls in the laboratory.
  • Valera supplied us with a new IR viewer that turns off automatically.
  • Steve is following up on all the action items identified at last week's safety audit. He has also compiled a detailed inventory of all lasers in the lab.
  • Steve obtained five more laser safety glasses with appropriate protection for both IR and green light (from the squeezer SHG). He is researching dual-wavelength glasses which have better visibility.
  • Bob's on vacation this week.

February 15, 2007

• There was a 40m Technical Advisory Committee (TAC) meeting on Feb 8th 8:30 Pacific. Rana showed these slides. The (draft) minutes from Ken Strain are here.
• There was a safety audit of the 40m lab on Feb 9, hosted by Rana, Steve and Bob. Several action items were identified and are being followed up on by Steve and Bob.
• IFO commissioning, Electronics, controls, computers:
  • Rob and Kirk have been working on getting the common mode servo working in the PRFPMI case with REFL 166 (ie, non-resonant sideband) signals. It's now commissioned & scripted, including the CM Boost stage.
  • Rob and Kirk have been working on improving the lock acquisition and control of the PRFPMI. It now locks reliably and robustly, and is fully scripted and (almost) fully automated.
  • Alex, Rob, Sam, Tobin and Valera began commissioning the new IFO dither-alignment system, using a new control computer (c1ass), front-end code from Rolf and Alex, and EPICS controls set up by Rob, Tobin and Valera. Valera returned to LLO and Tobin continues the commissioning work.
  • Rana set up some measurement scripts to measure the losses in the arms. He estimates a round-trip loss of 175 ppm (!) in the Y arm.
  • Rana asked Gari, Helena, and Bill Kells to help reduce the mirror losses by drag-wiping the test masses. We plan to do this some time in the spring.
  • Rana has been diagnosing problems in the dewhitening boards and coil drivers that are limiting our noise spectrum in the 100-1000 Hz regime, and is making plans to change the boards to reduce the noise.
  • Royal has been modifying the oplev whitening boards to reduce oplev servo noise at low frequencies.
  • We have HeNe oplev lasers on the ETMs, and they perform much better than the cheap diode lasers we were using before. Steve has a new layout of our BS oplev table (which sends oplev beams to the BS, PRM, and SRM) using a single new HeNe laser with his stable mounts, and will install it in the next week.
• IFO modeling:
  • Kirk has been running Optickle models of various MICH and PRC signals, with varying CARM offset, for undercoupled and overcoupled PRCs. Our PRC is currently quite lossy and undercoupled. Kirk is identifying useful MICH signals for different CARM offsets.
  • Kirk ran Bench sensitivity spectra for the 40m in PRFPMI configuration, with estimates of the (anomalously) large mirror loss. The shot noise limit should be a few times 1e-19 m/rtHz.
• DC detection:
  • According to the High Level Schedule in the wiki, DC Readout Phase A concludes on Valentine's Day, 2007. To celebrate, Rob took a valentine-themed DC readout-controlled DARM noise spectrum in the PRFPMI configuration, bottoming-out at 1e-18 m/rtHz in the 1-2kHz region.
• Vacuum squeezing:
  • Go and Osamu injected squeezed vacuum into the 40m in the signal-recycled-michelson (SRMI) configuration, scanned the sqeezing angle with a PZT mirror in the injection optics, and observed a -1.5 dB squeezed-vacuum-induced reduction in the noise at MHz frequencies (after subtracting electronics noise). This is a great achievement and a real milestone.
  • Go and Osamu have been working on noise-locking the squeeze angle of the injected squeezed vacuum, so that they can observe broadband squeezing at lower frequencies. The DC power level of the AS port beam in the SRMI configuration was too unstable and they couldn't noise lock. So Osamu fed the DC level to the Michelson offset to compensate for the fluctuation. This worked pretty well, and then they were able to noise-lock it. Its stability is not so good yet and is being optimized.
• Lab Infrastructure, Bake lab:

February 1, 2007

• There will be a 40m Technical Advisory Committee (TAC) meeting on Feb 8th 8:30 Pacific.
• IFO commissioning, Electronics, controls, computers:
  • Rob, Kirk and Rana have made lots of progress in exploring different paths towards full lock of the PRFPMI using different signals and handoffs. They are compiling a "map" of different paths to help understand what works and what doesn't.
  • Their most recent roughly-calibrated DARM noise spectrum gets down to ~3e-17 m/rtHz at ~ 1 kHz, and it is known that oplev and electronics noise will limit us at lower frequencies. Rana is working on improving the oplev filtering.
  • Kirk is modeling the PRFPMI with OpTickle to understand why lock is lost when the CARM offset is reduced to zero and the arms approach full power. He's learning that the conclusions depend sensitively on the assumed opticle losses, so he's working to get those right.
  • Rana got the 40m frame files mounted on the control computers and supplied matlab scripts to make it easy to display our data on demand.
  • Rob is reconfiguring conlog so that it doesn't log a huge amount of data unnecessarily. Pater Shawhan sent some advice on how to delete channels effectively.
  • Rob and Alex have started to commission the new alignment servo cpu (c1ass) and ASC software, in preparation for Valera's arrival to get the dither autoalignment system going.
• IFO modeling:
• DC detection:
  • Rob use the DC readout signal to control the DARM degree of freedom in a simple Fabry-Perot Michelson (FPMI) with CARM servo (both length and frequency paths). He sees a noise floor of ~ 1e-15 m/rtHz in the 1-4 kHz range.
• Vacuum squeezing:
  • Go and Osamu have we realigned and re-mode-matched the interferometer beam to the OPO cavity. They then measured the mode structure of the OPO cavity.
  • Osamu tweaked some optics and improved the stability of the SRMI lock.
  • Osamu and Go measured the noise floor in the SRMI configuration and determined that it was shot-noise dominated above 40 kHz.
  • Eugeniy arrived on Thursday, and he and Osamu and Go relaligned and re-mode-matched the pump field to the OPO cavity and got a parametric gain of about 10 back.
  • They are ready to inject squeezing Friday morning. They hope to see a 2 dB reduction in noise above 40 kHz.
• Lab Infrastructure, Bake lab:
  • Bob is preparing the 40m for a safety audit next week.
  • Bob has recieved an FTIR scanner for testing contaminants.

January 25, 2007

• January 24th, just 4pm, a new boy 'Yonosuke' was born at Verdugo Hospital in Glendale. Yonosuke means a child of the sun. Osamu and Hanako named him with a wish that he will be a bright and fair boy like California's sun. Baby's weight was 7.14lb, and both mother and baby are very fine.
• IFO commissioning, Electronics, controls, computers:
  • Tobin, Rob, and Kirk got back to smooth PRFPMI locking. Kirk did some modelling and came up with good signals for handoffs to smoothly go to the desired final configuration. They are implementing these on the 40m. They found and fixed some problems with demod phases. They got through the locking script and are ready to hand off to the SPOB166 signal, which requires a new SPOB166 photodiode. Work in progress.
  • Kirk measured the lengths of the two arm cavities by driving them through multiple fringes and observing the relative positions of the RF sideband peaks in the AS demod signals. The measurement suffers from an ambiguity as to which signal goes with the upper or lower sideband, and it's only really accurate to ~ 1 cm. But it's an interesting measurement, nonetheless. In the process, he observed a rather large 'length to angle' coupling in the YARM, which can be corrected.
  • Rob noticed that the conlog is recording data at rates exceeding 10MB per hour. We'll need to address this issue soon.
  • Steve and Osamu replaced the PRM oplev laser, and redid the layout of the table in preparation for putting in a HeNe laser.
  • Rob is cleaning up the multiple versions of front end code in the /cvs/cds code repository.
  • Larry brought us over a new op540m and swapped around some hard drives. op540m is now dual head and doesn't have a buggy OS.
• IFO modeling:
• DC detection:
  • Rob did a mode scan of the OMC transmitted beam by driving the length PZT with a triangle wave. He identified the modes visually on the CCD camera by deliberately misaligning the input beam. With the input beam aligned, The tallest peak in the spectrum is the TEM(m+n=2), which he thinks is due primarily to mode mismatch. Its height is 4% of the carrier maximum. He will tweak the output MMT.
• Vacuum squeezing:
  • Go measured the noise floor of the DC PD which is for squeezing-enhanced 40m measurements. He then Go estimated the shot noise level for various carrier power levels, assuming that the RF sidebands exiting the AS port when the SRMI is locked is a small fraction of the carrier. He figures that they could be shot-noise limited with 20-40 uW of carrier power.
  • Go and Osamu needed finer control of the picomotor that drives the in-vac squeezer pickoff mirror, which we normally control with a joystick. They set up some computer control to send commands to rotate the picomotor with the user-specified number of pulses.
• Lab Infrastructure, Bake lab:
  • Steve is studying the snow powder in Utah next week.

January 18, 2007

• IFO commissioning, Electronics, controls, computers:
  • Sam and Rob did some work to make SRC lock acquisition more flexible and robust.
  • Tobin and Kirk remeasured the length of the Mode Cleaner by driving MC_AO with a sine wave at 2 kHz, and scanned the 166 MHz marconi frequency while monitoring the power of the line at the AS port. The length seems to have increased by ~ 60 microns since last October. They adjusted the frequencies of all the RF signal generators accordingly.
  • Tobin mplemented a simple PID loop in Perl to control the C1:PSL-FSS_SLOWDC using C1:PSL-FSS_FAST as an error signal. This replaces some EPICS state code.
  • Kirk and Steve cleaned up the IPANG monitoring optics and got rid of some clipping.
  • Rolf and Alex installed new front-end CPU c1ass for dither-alignment of the main interferometer. The plan is for Valera to commission this system when he visits in February.
  • Alex built and installed new version of daqd in preparation to the addition of the new C1ASS into the DAQ. However, this produced errors when ezcademod was run, so Rana backed it out and informed Alex.
  • Justin has written a small perl program, called blinky, which provides a clean mechanism for making the PSLwatch, FSS slow servo, and autolockMCmain40m scripts blink their BEAT channels.
  • Justin tested all the IFO optic's dewhitening filters switching functionality using a battery of ifoTest scripts. All of the dewhitening filters appear to switch, as expected.
  • Justin wrote an MC test script, and started some documentation on all the IFO test scripts he is assembling.
• IFO modeling:
  • Kirk continues to use OpTickle to find a good signal for MICH control in SMOO locking of the PRFPMI.
• DC detection:
• Vacuum squeezing:
  • Osamu and Go tested the newly installed picomotor mirror for the first time after pumping down. It seems to work well, and they can direct the AS beam to the squeezing optics. However, the picomotor moves the mirror too fast to allow fine tuning; he and Osamu are working on slowing it down or servoing it.
  • Go updated he layout for the homodyne PD and some of the injection optics.
  • Osamu and Go then installed the mode-matching telescope and steering mirrors for the squeezing injection optics, and mode matched using light from the asymmetric port.
  • They then found good signals for locking the signal-recycled Michelson (SRMI) with a small MICH offset and only a small contribution from RF sidebands out the AS port, as measured with an optical spectrum analyzer at the AS port.
  • They then realigned the IFO beam to the OPO cavity using the steering mirrors. After the alignment, they installed and aligned the DC PD with high quantum efficiency. The PD is for measurements of squeezing-enhanced laser interferometry.
  • Go measured the noise floor of the DC PD used in his squeezing system, in preparation for measuring the shot noise of the LO field from the AS port in SRMI.
  • All the components are in place to perform a measurement using injected squeezed vacuum. They are gearing up for this now.
• Lab Infrastructure, Bake lab:
  • Tobin and Kirk snatched a discarded whiteboard and brought it to the 40m, so that we could "have a proper discussion on a whiteboard, like normal physicists."
  • Steve measured the particle counts at all the 40m flow benches, and measured 0 counts.
  • Steve discovered and fixed a leaking N2 supply line (used for control of the vacuum valves).
  • Steve notes that we now have several PZTs in vacuum and we need a hardware interlock to turn off the PZT HV when the vacuum pressure transits through the ionization region (either intentionally or because of some vacuum system failure). We are soliciting volunteers.

January 11, 2007

• IFO commissioning, Electronics, controls, computers:
  • Last Thursday, Steve vented the 40m (with no incident) in order to install the squeezer pickoff mirror and to search for clipping in the AS in-vac beamline.
  • Kirk, Osamu, Justin, Steve, and Rana started looking for the clipping. They locked the input MC in air and observed significant clipping on the TT-PZT1 sterring mirror in the BS chamber, just after the SRM. They also noted that the beam exiting the IFO is ~1.5 cm off from the center of SRM. This explains a lot: the AS beamline and TT-PZT mirrors were aligned using fiber-fed laser light from the OMC back to the IFO, under the assumption that the IFO beam is centered on the SRM. Since it isn't, the beam clips on TT-PZT1.
  • So, the crew moved the TT-PZT1 sterring mirror to eliminate the clipping there and further downstream (at the OMC MMT). Steve set up a jig to make a parallel transport of the mirror; it was translated parallel to the mirror surface by ~1 cm.
  • Further searching for clipping downstream of the TT-PZT1 sterring mirror did not yield anything significant. They then were able to lock the OMC.
  • Osamu and Go swapped out the old, nearly immobile squeezer pickoff mirror with a new rotating version, which works much better and swings into place in 1-2 seconds.
  • After some effort, they rebalanced the optical table (which is on a single-leg stack) with some aluminum ballast, and completed all the alignment of the various beams (to the squeezer, the OMC, and the in-air AS RF beamline). They checked that when the pickoff is swung out of the way, it does not clip the beams going to the OMC or to the AS RF beamline. Rana and Justin checked that the beam was well centered on all the in-vac optics.
  • After a final check by Steve, Rob, Osamu, Go, Rana, and Justin, including the installation of some beam dumps in strategic places, they sealed up the vacuum envelope and began pumpdown on Tuesday afternoon.
  • They left the envelope at 0.9 torr overnight, to see if any static charge buildup on the mirrors would discharge via the Paschen effect. Results were inconclusive, but it doesn't look like the mirrors changed much overnight.
  • On Wed morning, Steve turned on the turbopump and completed the pumpdown with no incident.
  • Within hours, the input MC, arms, Michelson, and OMC were re-aligned and locking up.
  • Kirk is adding extra zero/pole filters to all the LSC whitening board channels.
  • Rolf and Alex have installed a new computer, networking, and software for the new IFO alignment system, and commissioning will take place over the next few days.
  • Royal made some new cables for the ETMY oplev readout electronics to replace some inadequate ones.
  • Justin has been busily fixing and improving our scripts for maintenance and control of the interferometer, including: autoBURT, PSLwatch, trianglewave, LIGOenv.pm, LIGOio.pm, ifoTest, and PERL5LIB.
  • Justin noticed that BURT wasn't taking any snapshots because we had run out of subdirectories for it to use after 2006. There was no 2007 directory. So we have no BURT snapshots for the first week in 2007. Rana wrote a script to generate subdirectories up to the year 2200. Meanwhile, Justin fixed the autoburt script so that it would make the directories it needed, automatically.
  • Rana, Justin, and Kirk have been working on the smooth PRFPMI locking.
  • Tobin and Kirk measured and adjusted the FSS and MC loop UGFs and gains.
• IFO modeling:
• DC detection:
• Vacuum squeezing:
  • Now that the in-vacuum squeezer injection optics are installed, Go has designed the layout for the injection optics (including steering mirrors, faraday isolator, mode matching telescope, QPDs and DC PD), and the alignment procedure.
• Lab Infrastructure, Bake lab:
  • Larry and company installed a new network wireless transmitter in the lab with much more power, so we can hope to access medm on our laptops down the arms.

January 4, 2007

• IFO commissioning, Electronics, controls, computers:
  • Rana and Kirk worked on DRMI lock acquisition, and then Sam measured the locking time using his "run ten times, show the average acquire time" script. With some tweaking, he got it down to 2.5 seconds on average to lock.
  • Sam got the DRMI to lock on each of the lower and upper 166 MHz RF sidebands (forming optical spring or antispring), deterministically, by changing the REFL_166_Q demod phase.
  • Rana worked on Bram's autoalignment script, and it works pretty well for the xarm. One problem is that the input TT-PZT mirrors are too close together to make a tilt-free translation of the input beam. It would be a lot better if they were spaced further apart (~ 1 m).
  • We vented on Wednesday, with no problems, and opened up the output optic chamber. Rana locked the MC in air at the "wrong" polarization (to reduce the finesse and thus the stored power), to get light to the AS port. Go and Osamu will be installing the motorized mirror for the squeezing experiment, and Rana and company will be locating and eliminating the clipping that is observed on the AS beam, which is somewhere between TT-PZT1 and TT-PZT2.
• IFO modeling:
  • Kirk and Rana are using Optickle to understand the signals for the smooth locking of the PRFPMI. They have good error signals for CARM and DARM, and are looking for a good MICH DC locking signal (when locked to 0.8 of bright fringe).
• DC detection:
  • Before the holidays, Rob measured the resonances in the OMC TT-PZT sterring mirrors and found "a broad, fertile valley from 2-7 kHz where dither lines will be able to flourish." He locked 2 DOFs (both loops on the second tip/tilt: the angles) at relatively high bandwidth (50Hz), and this bandwidth can surely be pushed higher. Working on diagonalizing the sensing/actuation scheme.
  • Ben has finished the new DCPD schematic and pcb layout. It is being reviewed by Rob and Rana.
• Vacuum squeezing:
  • Before the holidays, Go and Osamu measured broadband squeezing while locking the OPO cavity with the subcarrier which was phase-locked to the MOPA laser and noise-locking the squeeze angle.
  • Osamu and Go locked a signal recycled Michelson (SRMI) for the squeezed vacuum injection, with greatly reduced sidebands exiting the AS port. They are looking for a good set of signals for controlling the MICH offset.
  • Shally implemented a new servo for phase-locking the second laser to the MOPA laser, and he and Go took the transfer functions of the fast and slow open loop gains.
  • Go is preparing a paper draft on the generation of a stable low-frequency squeezed vacuum field with periodically-poled KTP.
• Lab Infrastructure, Bake lab:
  • We ran out of N2 for the vacuum valves over Christmas, and the pressure rose to almost 4 mTorr. The PSL shutter interlock tripped, and the OMC PZT HV was turned off. Steve came in and restored vacuum mormal, shutters, and HV.

December 21, 2006

• IFO commissioning, Electronics, controls, computers:
  • Kirk, Valera and Sam locked the full IFO on double demod signals, and got part of the way through the common mode servo handoff.
  • Kirk is finding all kinds of problems with the LSC whitening board for our DC signals. He'll take some time to fix that board up, which could make a big difference in lock acquisition.
  • Osamu has taken free-swinging measurements of all the suspended optics and used Shihori's method to adjust the input matrices.
  • Rana upped the oplev loop gains on ETMX by a factor 10 to see if the suspensions had a problem with having an angular loop with UGF greater than their bounce/roll mode frequencies. The BW went from ~3 to 30 Hz. The noise below 30 Hz was suppressed and even the bounce and roll modes went down in the perror/yerror signals. There was "no drama".
  • Go and Osamu would like us to vent soon to install the flipper mirror for squeezed vacuum injection. We plan to do so in the first week of January. We will want to fix some clipping on the AS beam, if we can find out where it is happening. We may or may not be ready to also install new in-vac DC PD electronics. In the meantime, Osamu will work on developing a good method for locking the interferometer in a desired state with low transmitted RF power.
• IFO modeling:
• DC detection:
  • The front-end code for the output mode cleaner and alignment pzts has been updated to make the controls topology more flexible; the primary alteration has been the addition of ASC PZT control/excitation matrices. MEDM screens have been made/modified, foton file updated, testpoint parameter file updated, C0EDCU.ini updated, conlogger updated, autoBurt.req updated.
• Vacuum squeezing:
  • Osamu built a servo in a NIM module for phase-locking the frequency-shifted subcarrier laser to the main PSL laser. Go rebuilt a RMS detector in a NIM module for noise-locking the squeeze angle.
  • Go re-aligned the SHG and optimized the SHG conversion efficiency by adjusting the oven temp.
  • Go reoptimized the OPO parametric gain by adjusting the crystal temperature.
  • Go reoptimized the homodyne efficiency by realigning and mode-matching the LO field, and got a homodyne visibility of nearly 100%.
• Lab Infrastructure, Bake lab:
  • Steve made plots of the 180-day history of the PSL, showing that the Mach Zehnder and FSS are much more stable after recent work.

December 14, 2006

• IFO commissioning, Electronics, controls, computers:
  • There was more work on smooth locking of the PRFPMI, by Valera, Rob, Sam, Rana and Kirk.
  • There was more work on the common mode servo, by Rana, Sam and Kirk. An offset that was pulling the MC off resonance and out of alignment was found and trimmed away. After that, they were able to lock the YARM using the CM servo board, with an 18 kHz UGF with 30 degrees of phase. Sam also wrote a detailed note on the common mode servo topology.
  • Sam, Rana, and Valera set the LSC Output Matrix today to better 'diagonalize' the DRMI. The new settings make it so that driving MICH doesn't change the physical PRC and SRC lengths. Driving the SRC loop also doesn't change the PRC loop. This had a remarkably good influence on the MICH loop, removing an extra unity gain crossing that it had.
  • Valera, Rana, and Bram centered the beam spot on the ETMX by minimizing the angle-to-length coupling in the x-arm. They then centered the transmitted beam of the X-arm on to the QPDX.
  • Valera, Bram, Rob and Rana have come up with a new alignment system for the 40m making use of dithering of the suspended optics and steering PZTs in pitch and yaw. We decided that the best way to implement this new system was not to modify our existing ASC system (which doesn't do much but also doesn't have much CPU power) but instead to assemble a new PCIX-based system. This will also give us room to grow and implement more control systems, and/or migrate the existing SUS, LSC, etc systems to the new architecture.
  • Rolf is starting work on the controls for this new alignment system. This will require a computer, I/O chassis, ADC card and some networking equipment to be purchased. We will use our Matlab to CDS code generation tool for the software, which will require us to add a few more parts to our library. Parts will include a reflective memory comms part to send/rcv signals to/from the existing LSC and ASC computers and a new matrix part which can be sized by the user (presently support up to 8x8, but new system requires much larger, so we'll make that part such that the user can specify any size). We have all the hardware in house (except the network switch), but we will need to purchase replacements. Goal is to have this installed by mid January.
• IFO modeling:
  • Osamu and Kentaro have developed a new RF modulation scheme for control of AdLIGO, based on lower RF frequencies (4.5 and 58.5 MHz), which met with favor at last week's AdLIGO ISC summit at MIT. Continued development of the scheme and the modelling is in progress.
  • Kirk fixed a bug in the way the Bench quantum noise model handled losses, and updated the predictions for the 40m quantum noise curve.
• DC detection:
  • Ben is doing a quick re-design of the DCPD head electronics to allow for internal ribbon cables to minimize harming the pin seals, and to install another stage of whitening. We'll install this in a new nipple and vent sometime soon to swap out the old and swap in this new DCPD electronics.
  • The krypton leaking out of the nipple housing the DC photodiode preamp is almost gone. At day 2 after pumpdown, the peak at AMU=84 was at 6e-8Torr partial pressure. Now, after 67 days of pumping with maglev only the partial pressure is 2e-11Torr.
• Vacuum squeezing:
  • Go installed a Lightwave 126 laser from MIT to use in creating a frequency-shifted subcarrier. He and Osamu spent some time tuning the temperature control of both this laser and our main MOPA, to find the beat between them. With advice from Thomas at MIT, they finally found the beat. Go then realigned and mode-matched the new 126 subcarrier to the OPO cavity in the TEM00 modes which are orthogonally polarized to the seed field. With the 126's sufficient laser power, he can lock the cavity stably now. He then phase-locked the second laser to the MOPA laser with its fast actuator.
  • Go has now readied almost everything in his squeezer on the PSL table and further progress requires light to/from the main interferometer. We will need to vent soon to install his improved flipper mirror which can be flipped into place to route the asymmetric port beam to the squeezer as opposed to the DC readout system. He and Osamu will develop a detailed plan for locking the interferometer in a desired configuration (simplest would be a signal-recylced Michelson) where they are shot-noise limited above ~ 30 kHz, and take a series of measurements with the injected squeezed vacuum.
• Lab Infrastructure, Bake lab:
  • Bob has been working with visitors on the assembly of magnets for a new AdLIGO Faraday isolator. He will bake the completed assembly over the weekend.

December 7, 2006

• DC detection:
  • After last week's major milestone (commissioning the length and alignment control servos of the OMC and using the DC readout to control the michelson and measure noise spectra of the michelson and the smoothly-locked offset PRFPMI), Rob's attention has turned to lock acquisition work.
• Vacuum squeezing:
  • Go phase-locked the Lightwave 120 laser to the MOPA laser with a frequency shift of 610MHz. He used this frequency-shifted subcarrier to lock the OPO.
• IFO commissioning, Electronics, controls, computers:
  • Bram and Rana worked on developing a DC alignment procedure by dithering mirrors at frequencies of 7 - 11 Hz.
  • Rob, Valera, Rana and Kirk worked on hand off from where the smooth PRFPMI script gets us to - i.e. arms and Mich locked to 1/2 fringe using DC signals, and PRC on resonance locked using REFL_33I (SRM is misaligned) - to a full RF signal scheme. They encountered a variety of problems, but are making progress.
  • Kirk aligned the OSA on the PSL table and measured the modulation depth of the 33MHz and 166MHz signals after the Mach-Zender.
  • Sam, Valera, Rob and Kirk worked on the implementation of the common mode servo, using YARM control as a pathfinder. Sam has written up a document describing the implementation, which is very similar to what is done at the sites.
  • Rob noticed that the MC was having trouble locking. After some sleuthing, he concluded that the cables in the drive chain for MC suspensions get loose in response to seismic vibrations, and require reseating. After re-seating all the MC suspension cables, the MC was locking again. Rob suspects that this problem affects the other suspended optics as well.
  • Rob, Valera and Bram are developing a new global scheme for DC alignment using low-frequency dithering of the suspended optics. The idea is to implement the digital dither-demod scheme used for OMC control, in our IFO ASC system. This will require some help from Rolf. Rob modified the LSC code to send signals to the ASC processor, in anticipation of this.
• IFO modeling:
  • Osamu is at MIT for a summit on AdLIGO ISC.
• Lab Infrastructure, Bake lab:
  • Bob continues to tend to the care and feeding of the 40m lab while Steve is on vacation.
  • Bob had a turbo pump on oven C fail in the middle of a bake job. He replaced it before the oven temp got too high.

November 30, 2006

• DC detection:
  • Rob has commissioned the length and alignment control systems for the output mode cleaner. Using dithers in the kHz range, the OMC length control locks with UGF ~ 300 Hz. One by one, all 4 ASC loops lock and run with UGF ~ 25 Hz. However, there was considerable coupling between pitch and yaw. This seems to be due to the fact that the dither frequency is above the first resonance of the PZT mirror mounts (~ 600 Hz), so the whole thing is shaking. Rob then went down to several hundred Hz dither lines, and can now lock all 4 degrees of freedom simultaneously and robustly, with UGFs around 7 Hz.
  • Rob used the reflected light to measure the visibility of the OMC. With no attempt at mode-matching (beyond what was done during the installation vent), he gets OMC visibility > 92.5%.
  • Working with the main IFO at Michelson mid-fringe, Rob transfered MICH control from the RF signal to the DC signal; it worked the first time.
  • With MICH under DC control, Rob measured noise spectrum and calibrated it using the technique Kirk and Rana used to calibrate the Michelson last month (the level is ~ 1e-11 m/rtHz at 100 Hz falling to ~ 1e-12 at a few kHz). The comparison of the measured noise spectrum shows that the DC control has far less dark (electronics) noise and 60 Hz harmonics, and is quieter than the RF readout above ~ 150 Hz (and about the same, below that).
  • Although there's plenty more optimization and measurements to be done (including control of the PRFPMI), the above achievements meet our initial goals for this first DC readout prototype. This is a significant milestone!
  • Rana swept the OMC length and measured the higher order mode structure transmitted through the OPD with poorly-aligned light from the PRFPMI. We can use the higher order mode peak locations and the distance between the TEM00 modes to determine the g-factor and the curved mirror's radius.
  • Rob found that the DC photodiode whitening filters in the in-vac preamp are not quite right, and the digital compensation filters are not compensating. He modified the digital filters, but we need more whitening for low noise running with the full IFO. Rob is working with Ben to prepare a modified preamp to be installed at our next vent, along with a new preamp can / nipple to replace the one that is leaking due to mechanical stresses during assembly.
  • We may vent in the next few weeks to replace the DC photodiode preamp / nipple and also Go's movable mirror for vacuum squeezing.
  • Rob is scriptifying the DC readout controls, to automate the locking and re-locking of the length and alignment servos.
• Vacuum squeezing:
  • Go, Eugeniy and Shally re-established ~6 dB of squeezing after much optimization of the squeezer system. They noise-locked the OPO and got 3db of squeezing broadband.
  • Go and Osamu measured the PSL laser noise to see if it was shot-noise limited, and determined that it will be difficult to see the effect of squeezing below ~ 10 kHz.
  • Go's flipper mechanism for inserting a mirror to inject squeezed vacuum into the interferometer works well and is now ready for installation into the vacuum system. The ouptut optic chamber will most likely need rebalancing, so ballast is being prepared.
  • Osamu tested a new photodiode for squeezing monitoring; it looks good.
  • Go tried using an AOM to generate a 600 MHz single sideband for OPO locking, but it didn't work. He's going back to using an auxilliary laser, a Lightwave 120. The throughput is low and the control signal has an SNR of ~5, but the OPO cavity locks and for now he will live with it. It's less noisy than noise locking, anyway. Now he's working on locking the auxilliary laser to the PSL MOPA frequency.
  • Go's next step is to set up the squeezing phase locking, and then he's just about ready to inject squeezed vacuum into the IFO. We will install his mirror in the coming weeks. Go plans to have a meeting of interested parties to plan this next phase of the squeezing experiment.
  • Go gave a LIGO seminar last week on applications of servo-controlled optical cavities in biomedical imaging.
  • Kirk gave a CaJAGWR talk on Tuesday reviewing work on quantum squeezing for GW interferometry.
• IFO commissioning, Electronics, controls, computers:
  • Kirk, Sam and Rana worked on the new common mode servo and tested it by locking the YARM, raising the UGF from 300 Hz to somewhere in the neighborhood of 6 kHz. They had some problems in the CM handoff, traced to a bad XYCOM board that was replaced.
  • Kirk and Rana have been working on scripts for locking the PRFPMI smoothly. The scripts work and the PRFPMI can be locked smoothly and brought the arms to half fringe, the Michelson at half fringe, and the PRC fully resonant. A bit more work is needed to switch the arms over to RF signals, switch on the CM servo, and bring the PRFPMI to full power.
  • Kirk and Rana locked the IFO in the FPMI configuration (without PR) and then made the transition from locking the arms at 1/2 fringe with the DC signal, to full resonance using RF signals.
  • Bram has been designing slow alignment servo loops (Perl scripts interfacing with EPICS channels) to control the position of the beam on the ITMs and ETMs or the arm cavities (to minimize angle -> length couplings) and the launch positions/angles of the beam into the arms (to maximize coupling into the arms). He has a document and wiki page describing the system, and is commissioning his scripts.
  • Rana did some work on all the oplev servos, and got the UGF up to 5 Hz on all of them.
• IFO modeling:
  • Osamu has released a new version of his loopnoise matlab package, which calculates the noise contribution from IFO control loops to the GW channel in AdLIGO. He concludes that the coupling from the SRC loop to DARM is rather large. Quantum noise due from the RF vacuum is important (especially at low frequencies) and he's been working on including it, but it is difficult. He's off to MIT next week for a summit on AdLIGO ISC issues.
  • Kirk found and fixed an error in the way the BENCH code implements non-100% quantum efficiency in the quantum noise. He will update the official BENCH code repository in the AdLIGO Wiki, soon.
• Lab Infrastructure, Bake lab:
  • Luke Williams will be visiting from UFl and will be assembling big magnets in the Bake Lab clean room. Bob is preparing the lab (removing all magnetically susceptible materials). He has received the magnets; they are indeed quite powerful.
  • Bob continues to tend to the care and feeding of the 40m lab while Steve is on vacation.

November 16, 2006

• Lots of good progress this week, on tuning the PSL, MZ, and mode cleaner; lock acquisition of the dual-recycled Fabry-Perot michelson; commissioning of the DC readout servos; and vacuum squeezing.
• DC detection:
  • Rob revamped the front end controls and the MEDM screens for the DC readout system. He then deleted the old OMC channels from the conlog database and the DAQ, and added the new ones; and added an autoBurt.req file.
  • Rob digitally locked the OMC length with a ugf of 200Hz and a dither frequency of 12kHz. He then dithered a tip-tilt PZT at 9kHz in pitch and 10kHz in yaw, saw reasonable error signals, implemented a servo filter, and locked the two degrees of freedom. He easily got a UGF of 80 Hz. The other tip-tilt PZT will come next.
  • Rob got a camera onto the OMC transmitted beam (it looks nice and round) and a thorlabs PD onto the OMC reflected beam and thence into the digital control system.
• Vacuum squeezing:
  • After several weeks of optimization work, Eugeniy, Shally, and Go observed about 4.5 dB of squeezing and about 8.0 dB of anti-squeezing 900kHz before subtracting the electronic noise, and 6 dB after subtracting electronic noise. They also noise-locked the homodyne angle while locking the OPO cavity manually.
  • Go and Eugeniy tried without success to use two different auxiliary lasers to generate a 610MHz subcarrier subcarrier to be used for locking the OPO cavity. They are now turning to an AOM to generate it from the main beam.
• IFO commissioning, Electronics, controls, computers:
  • After all the good work on the PSL / MZ / MC (described below), Bram, Kirk, Rana, and Sam locked up the IFO (with the arms at half-resonance) with no problems. Handoff to the double-demod signals and DARM-RF was no problem. They swept the MICH, PRC, SRC and CARM loops and optimized the gains. They successfully handed off CARM from driving the differential ETMs to driving the MC_L path. Next step is to get the common mode servo working well and then start reducing the CARM offset to get back to full RSE.
  • Kirk, Rana, and Sam are configuring the common mode servo board to match the topology employed at the sites. The input comes from POY_33_I, a digitized slow output goes to MC_L, and an analog fast output drives the mode cleaner "additive offset" for the frequency stabilization loop. They used this to lock the y arm. They also implemented a servo script to offload MC_F values to the ETMs to keep the VCO from railing.
  • Kirk and Rana found that the PSL FSS beam was only singly passed through the AOM instead of double-passed, so they realigned the entire FSS/RefCav/AOM path on the PSL table. This recovered a bunch of phase in the MCF servo path, allowing them to increase the MC servo UGF to 50 kHz.
  • Rana, Sam, Rob and Kirk reworked the Mach Zehnder servo board, allowing them to increase the gain at DC by a factor 100.
  • Bram replaced the ETMX oplev diode laser with a HeNe laser and reduced the number of bounces, as he did with ETMY before. He had to flip the sign of the servo gain to get it to work.
  • Steve designed the optical mounting hardware for our new oplev He/Ne lasers. This is an effort to reduce cost and improve performance with optical levels. It will be ready in three weeks.
  • Royal modified and re-installed the whitening board on the ETMX to give it the same analog and digital filters that she added to the ETMY, which significantly reduced the noise. She also modified the digital filters by adding two new dewhitening filters for each quadrant to deamplify the signal.
  • Rana wrote and employed scripts to optimize the the demod phase for POX and POY with single arm lock. The optimized phases were very different from what was there before!
  • Rana turned off the MC2 oplev and guided the beam transmitted through MC2 into the old MC2 oplev QPD.
  • Rana locked the DRMI using single-demod REFL RF signals, then tuned the double-demod phases. Now the handoff from single-demod to double-demod is smooth.
  • Rana and Kirk locked the michelson and took a calibrated spectrum. The noise is huge.
  • Rana found evidence of excess noise in the OSEMs of several suspensions, probably electronic. To be investigated.
  • Kirk and Sam measured and tweaked the PSL/IOO servos, increasing their UGFs to 580 kHz for the FSS and 65 kHz for the MC. The crossover frequency from MCL to MCF is now at 85 Hz. These servos are now working very well.
  • Steve reinstalled beam tube guards on the PSL table so that they do not clip the ISS pickoff.
  • Steve noticed low beam centering on the second mode matching lens into the MC. He and Rana rebuilt the mount and the centering is good now. The pointing is recovered, however the MCT power went down a little bit. This needs more fine tuning. The psl_pos_qpd was set up again as preparation for this work. The psl_ang_qpd does not work. The 35.5MHz EOM is glowing more than it should, requiring better alignment.
  • After this, Rob, Kirk, and Bram recentered the MCR beam on the WFSs and the MC locked happily.
• IFO modeling:
  • Rana and Kirk assembled a Bench calculation of the expected noise spectrum from the 40m PR-FPMI. They found and fixed a problem in the calculation of the radiation pressure noise.
• Lab Infrastructure, Bake lab:
  • Steve noted that the leybold-secu valve must have cracked o-ring seal (it is 30y old); Bob will take apart and fix it.
  • MIT's Spectra Physics argon ion laser was handed over to Rod Luna. Model 2080-25S & 2580C, SN 298&629. D. Shoemaker wanted the MIT property tag but Steve couldn't find it.
  • Bob will tend to the care and feeding of the 40m lab while Steve is on vacation for three weeks.

November 9, 2006

• DC detection:
  • Rob has been working on the DC readout control software. He is using Rolf's simulink model to modify the logic and flow, and auto-rebuild the front end software. The build was successful and is running, but testing was hampered by malfunctioning testpoints. Alex came in and showed Rob how to fix the testpoints, so tests of the front end software are now in progress.
  • Rob is building a raft of snazzy new MEDM screens for control of the DC readout front end.
• Vacuum squeezing:
  • Go and Osamu assembled the homodyne photodiodes and they tested fine. Go installed a new homodyne detector placed on the PSL table right before injection, to check the level of squeezing.
  • Eugeniy is here from MIT. He and Go installed the homodyne detectors and re-optimized the homodyne visibility by carefully balancing the splitting ratio of the homodyne BS. It's close to 99% now.
  • Eugeniy and Go set up a Lightwave 120 laser and associated optics for a 610MHz subcarrier to be used for locking the OPO cavity.
  • Go's new mirror translation stage parts have arrived and are being baked. If the picomotor installed in the vacuum chamber is really broken, we will ready a spare for installation the next time we vent.
  • Go and Osamu measured the laser intensity noise at intensities and frequencies of interest for squeezing, with ISS turned on. They established that the laser is shot-noise limited above 30 kHz (where the noise spectrum scales with sqrt(power)).
• IFO commissioning, Electronics, controls, computers:
  • Kirk and Sam have been working on the MC servo. The OLTF has a gain peak at 100 kHz which prevents us from going to higher gain. They believe the problem is due to some loss of phase in the FSS loop, so they will work on realigning that (including the beam path through the AOM). Once they tweak up FSS alignment, they will return to increasing the gain on the mode cleaner.
  • Kirk reports that after all the work done on the Mach Zehnder, it now works well, is much more stable, with 1000x more gain at low frequency. However, there are still problems with a flaky gain slider; under investigation.
  • Sam is working with Kirk, Rob, and Rana to get the common mode servo working. The input signals and output actuators are still to be defined.
  • There was an unexplained change in the MC RFPD DC level. Steve and Rana took some power measurements around the PSL table. Suspicion fell on the MC RFPD itself. Rana pulled out the box and found multiple problems, which he addressed. He also noted several pits on the PD surface. He figures we can live with it for now, so long as we put the beam un an undamaged part of the surface.
  • Rana reports that the ETMY oplev, which has a HeNe laser instead of the cheap diode lasers on all the other suspensed optics, is much more stable than the others. He and Bram improved the mounting of that laser; the noise spectrum is significantly improved. Steve has ordered more clamp mounts for more lasers, to be installed on all the test masses and the BS. Steve and Rana also cleaned up the layout of the ITMX oplev.
  • After all this PSL, Mach Zehnder, and mode cleaner work, Rana and company plan to return to lock acquisition work in earnest, in preparation for the DC readout measurements.
  • Bram is working on scripts to autoalign the IFO mirrors and to facilitate manual alignment. His scripts seem to have a Star Wars theme, for some reason.
• IFO modeling:
  • Osamu continues to develop the Optickle-based modeling code for AdLIGO, including quantum noise from all important input ports. He is working on the addition of quantum noise to the RF sidebands used to sense the auxilliary degrees of freedom, in order to correctly model their contribution to the noise in the GW channel.
• Lab Infrastructure, Bake lab:
  • Our OMC photodiode preamp in-vac nipple is leaking Krypton. Steve reports that the 2e-8Torr amu 84 amplitude droped to 2e-9 Torr in 16 days; the doubly ionized amu 42 fell the same rate. The hydrocarbon level is normal, not rising .... no action needed.
  • However, Steve points out that we still need a vacuum valve V1 interlock in case of power failures.
  • Roughing pump RP2 did not work for the last pumpdown. Steve is working on the valve for the the old Leybold pump. A small drypump was send back to Varian for rebuild.
  • The AS ISC table enclosure connection to the output optic chamber needs some modification to accommodate the OMC exiting beams. Steve has the parts in the machine shop.
  • Steve moved the OMC alignment fiber to the top of the PSL enclosure, allowing us to once again close the doors completely.
  • Steve continues to monitor the PSL, Mach Zehnder, mode cleaner and IOO power and other signals to understand the slow drifts. There are so many things evolving in the lab that it is hard to compare performance.
  • Steve continues to monitor the power supplies for the SUS and other systems, and note any changes larger than 0.2A.
  • Bob has been ordering parts and supplies for various activities around the lab.

November 2, 2006

• DC detection:
  • Rob and company are commissioning the DC readout system and associated controls. To lowest order, everything is roughly working: The OMC PZT is scanning through transmission resonances. The OMC transmitted light is picked off and directed to a camera on the AS-ISCT table. The tip-tilt PZTs are biased and have been swept. Slow channels (136 of them) are now in frames. 1391 channels are added to conlog (which now monitors 35,000 channels total). Testpoints at 32K are in DTT, not yet in frames (soon). DC PD transmitted light is available in RFM to the main interferometer LSC system. Rob has modified the lock acquisition front end code to add these signals. Bram and Rob are working on locking the OMC with the new digital controls. Some additional front end code is required; in progress. Next: the dither-locking of the tip-tilt PZT alignment. At that point, it becomes meaningful to characterize the OMC performance.
• Vacuum squeezing:
  • Go used a cylindrical lens in a newly designed mode matching telescope to correct the beam ellipticity of the OPO output beam. He measured the beam profile; it is now nicely circular.
  • Go then designed a mode-matching telescope for injecting the squeezed vacuum field, and a a new mode-matching beam expander design for the LO field.
  • Osamu is testing the squeezer homodyne photodetector.
  • Go performed a Noise Budget Analysis of the photodetector for Squeezing-Enhanced SRMI, and determined that a different opamp will reduce the electronics noise relative to the shot noise in the relevant frequency band.
• IFO commissioning, Electronics, controls, computers:
  • There was a power outage Monday early AM, discovered by 40m staff at 9am. Most everything was back up by noon; we're getting better at this! A couple of bad cables were found in the course of recovering the many systems.
  • Rana is renewing efforts to re-establish full lock with full power. He has locked the arms and the michelson and the PRM, but there are some signals missing after the vent that need to be re-established.
  • Bram is working on the scripting required to auto-align the full IFO via dithering of all the suspended optics. At present, this requires the arbitrary waveform generator (AWG) which tends to saturate the network when more than a couple of signals are applied. We need Rolf's new code (for DC readout) applied to our (old) SUS/ASC system, so that we can dither the alignment degrees of freedom without AWG. In the meantime, Bram will use AWG.
  • Rob has been working on getting scripts working on our linux machines as well as solaris. We still have a variety of problems with our solaris machines as well.
  • Sam and Kirk remeasured the input mode cleaner length yet again, and updated the RF signal generators appropriately.
  • Kirk and Sam installed the CM servo board and all connections are checked out. They haven't tested it at OLTF level or using it to control IFO; that comes next. They are assembling a test plan and putting documentation in the wiki.
  • Kirk is working on optimizing the Mach Zehnder servo, improving the EPICS control screens, and is monitoring and characterizing its stability over days. After the MZ changes, the mode cleaner demod phase needed adjustment; they added 30 cm of cable to the MC RF PD path.
  • Kirk is setting up systems to monitor the AM on the RF sidebands transmitted through the input mode cleaner.
  • Rob, Steve, and Kirk realigned the MC onto the WFS detectors and CCD camera.
  • Bob got spare 400 GB hard disk spares for our DAQ RAID array.
• IFO modeling:
  • Osamu continues to develop the Optickle-based modeling code, including quantum noise from all important input ports, as part of the exploration of alternative length sensing schemes for AdLIGO.
  • Bram and Sam are working on the AdLIGO OMC layout.
• Lab Infrastructure, Bake lab:

October 26, 2006

• DC detection:
  • Ben and Rich brought over their PZT drivers and the QPD whitening electronics (with documentation!), and installed them in the DC readout controls rack (1X5).
  • Ben and Bob ran conduit to power up the DC readout control rack, including HV for the PZT drivers. Everything is now powered up and, to lowest order, working.
  • Rob used the new system to put DC bias on the tip-tilt PZTs and they seem to work. He's beginning to test the OMC PZT controls.
  • We will place two QPDs on the AS table to monitor the OMC reflected beam exiting the chamber.
  • Alex worked some more on getting dtt/dataviewer running with our new 32 kHz system.
  • Steve reports that the Krypton peak from the leaky DC photodetector nipple is going down; so the leak is bigger than we estimated.
• Vacuum squeezing:
  • Go has designed a new propeller-type mirror mount that can be rotated by the picomotor. parts are bring machined in the shop and a new picomotor has been ordered.
  • Go measured the OPO mode structure in both polarizations. The OPO output beam is very elliptical; Go is designing a new mode-matching solution including a cylindrical lens.
  • Go and Osamu are setting up setting up subcarrier generation optics to lock the OPO to a frequency shifted subcarrier.
  • Go and Osamu are installing Riccardo's photodiode and beginning tests.
  • Kirk and Go installed some of the squeezed vacuum injection optics.
• IFO commissioning, Electronics, controls, computers:
  • Sam and Kirk are preparing to replace our old CM board with a new style board at the LSC rack. They are in the process of setting up all the connections and defining the cross-connect work, and readying all the EPICS database files. They will make sure the LSC rack drawings get updated.
  • After Kirk replaced the mounts of the EOMs in the Mach Zehnder last week with more stable ones, we expected the MZ transmitted beam to get more stable. But the weather started changing at around the same time, and the laser power started to slowly vary. (Not sure if this is real or due to monitoring photodiodes being strongly humidity-dependent). Steve put a calorimeter in a portion of the PSL beam to monitor it. Kirk will keep a close eye on the MZ error signal and transmitted power.
  • Sam and Kirk measured the mode cleaner length, and found it to be 1.4 mm longer than the last measurement (2.5 months ago). They have adjusted the RF feequencies accordingly.
  • Rob got the TDS suite of software tools, including ezlockin, to work with the new install of GDS (on Solaris), and after much effort, also on our linux controls computers. He cleaned up much of the controls software, but there is still much work to be done to make all controls computers useful for all tasks.
  • Royal added whitening filters to a Pentek Generic Interface board, and installed it at the ETMY optical lever.
• IFO modeling:
  • Osamu is making much progress in the modeling of AdLIGO length control using OpTickle. He has added radiation pressure to the simulation, and has generated quantum noise curves. He is exploring the 9/45 MHz RF modulation scheme as well as the nominal 9/180 MHz scheme. He will present results at an upcoming AdLIGO meeting.
• Lab Infrastructure, Bake lab:

October 19, 2006

• The 40m Technical Advisory Committee (40m TAC) met by telecon this morning (10/19/06). Rana led the discussion, reviewing recent progress and future near-term and long-term plans. Details are in his slides and the Wiki.
• The Honorable Jack McConnell, First Minister of Scotland, visited the 40m on Monday, with an entourage from the British Consulate. Special guests included Jim Hough and Shiela Rowen of U Glasgow / LIGO.
• DC detection:
  • Valera, Rana and Sam put the IFO into a bright Michelson, worked on the alignment using the BS and ITMs, and eventually saw nice symmetric fringes on the OMC, of about 75% of the input power. They then used the REFL PD to dither lock the cavity, using a 12 kHz dither and the SRS lock in amplifier. Rob found the OMC transmitted beam and got it onto a camera; it looks round.
  • Alex has been working on getting all of the software tools working with the new DC readout controls. The controls run at 32KHz, above our previous highest rep rates of 16KHz, and are not VME and reflected memory network based, so some Framebuilder and DTT codes still need modifications to support this. Alex has installed a new version of the software on the controls computer (c1omc). He also updated dtt (to work with 32KHz data rate?) and moved the test point manager to another machine.
  • Rich finished testing on the OMC Length Driver Chassis. All functions are verified to be within design specifications. He has also finished testing on the tip/tilt Piezo driver chassis used to drive the Piezo Jena piezo steering units.
  • Ben has tested the QPD Whitening Interface board, and it is now installed in its box, and the front and back panels are on. It powers up fine, and will be installed soon.
  • Rob has been refining the DC readout EPICS controls screens and learning how to use them.
  • Bram, Bob and Ben are wiring up the low voltage and high voltage power supplies for the DC readout controls electronics.
  • Bottom line: we should have all the hardware and software in place, and begin commissioning the new DC readout controls, next week.
• Vacuum squeezing:
  • Go subtracted the electric noise from the shot noise and squeezed noise, and found about 5-6 dB of squeezing and about 12-13 dB of anti-squeezing when the squeeze angle was scanned. The imbalance between the two levels is mostly due to the detection efficiency.
  • The in-vac picomotor translation stage and mirror mount for the squeezer experiment is not working. Go has designed a new propeller-type mirror mount that can be rotated by the picomotor. We will vent sometime in the future to fix this system.
• IFO commissioning, Electronics, controls, computers:
  • Rana has been adjusting the MC mirrors to find some compromise between centering the MC resonant beam on the optics and getting the transmitted beam through the faraday isolator without clipping. Ongoing work.
  • Rana found large offsets on the input MC WFS DC outputs, so he reset the appropriate database entries, recented the beams, restarted the system, adjusted the gain, and all seems well.
  • Rob re-established lock of the dual-recycled Michelson. He sees clear evidence for clipping, maybe in the Faraday isolator.
  • Kirk, with help from Rob, Rana, Sam and Bram, replaced the makeshift Pockels cell mounts on the Mach Zehnder with tip/tilt stages mounted on anodised aluminium mounts. They decided to take out the 29MHz modulator altogether, leaving the 33MHz PM in one arm of the MZ and the 166MHz PM in the other. The 33MHz and 29MHz signals are now summed electronically, and fed to the 33MHz modulator. Kirk realigned everything and adjusted the servo loop offset to get exactly onto the bright fringe, minimizing AM. The fringe visibility went way up, the error signal is now much more stable, and there's 5% more power into the MC.
  • Steve and Kirk noticed that the laser power was dropping, as was the power transmitted through the input mode cleaner. Rob found the problem: the c1psl processor was down, for some unknown reason. He rebooted it and relocked everything, but problems with dataviewer (related to the DC readout electronics) prevented him from verifying that the PSL was restored to the right state.
• IFO modeling:
• Lab Infrastructure, Bake lab:

October 12, 2006

• We have two new Aussie visitors, Kirk McKenzie and Bram Slagmolen. Please make them welcome! Steve gave them a safety walkthrough, and they have had their entrance eye exam.
• Final reports and talks from our five summer SURF/REU students are now posted on the 40m web site.
• DC detection:
  • We ended our vent last Thursday and Steve began the pumpdown Thursday morning. The pumpdown proceeded without incident, was complete by the end of the day, and the mode cleaner was locked by Thursday evening.
  • After the vent, Rob, Sam and Valera reestablished MC lock and lock of both arms. With the michelson on the bright fringe, they sent light into the OMC and saw fringes as well as signals on the DC photodiodes. So, to very lowest order, the DC readout beamline is operative.
  • To take the next step, we need the digital controls for the OMC and steering PZTs. Ben and Rich report that the remaining electronics modules (the OMC and tip-tilt PZT driver boards and the QPD whitening board) should be ready next week. Commissioning will then begin in earnest. Rob has written up a commissioning plan in the Wiki.
  • Alex has plugged in the DC readout controller computer into the reflective memory loop, configured data acquisition channels for it, and the new frame builder code was started. Still to do: install 32KHz DTT, dataviewer software. However, these changes seem to have broken ezlockin, which uses NDS.
  • Ben's testing on the QPD Whitening Interface board went fine. He is going to use the results to write the test procedure for future boards.
  • The front and back panels for the QPD Whitening chassis should be here by the end of this week, whereupon Ben will assemble and deliver the unit to the 40m.
  • Rich received all boards back from the stuffing house for the OMC PZT driver and Tip/Tilt driver. Have tested most of the OMC length board and it performs well at 300 volts peak-to-peak so far. Still need to test the more complex Tip/Tilt board.
  • Bob has been building and running cables for the DC readout controls and mounting equipment in racks.
  • Our DC photodiode electronics nipple is leaking. We had filled it with krypton, which shows up clearly in the RGA. The preamp PCB is supported mechanically by the vacuum feedthroughs, and we figure it put too much stress on the feedthroughs, causing the leak. Valera calculates that it will take around a year for all the krypton will leak out; we'll fix it long before then. Bob has ordered a new vacuum nipple and feedthroughs, and we are thinking about a redesign of the DC PD mount and in-vac electronics.
• Vacuum squeezing:
  • Go and Osamu have recalculated the predicted quantum noise floor for the dual-recycled Michelson, with and without squeezing.
  • The picomotor controlling the translation stage that inserts the squeezed vacuum into the interferometer is not behaving well; it seems to be whirring and it's shaking the MC, so the cables seem ok, but the stage is barely moving. Most likely, it is broken and will need to be fixed the next time we vent.
  • Go, Kirk and Nergis are working on the squeezing test plan, and are setting up a Wiki for the experiment.
• IFO commissioning, Electronics, controls, computers:
  • The OSEMs for the MC and other suspended optics were adjusted during the vent. Steve measured the inductance of all the coils, they all look fine. The sensors are all roughly in the middle of their range. so now Rana, Rob and Bram are adjusting the coil driver gains to minimize angle-to-length couplings.
  • The whole gang (Sam, Bram, Rob, Kirk, Valera, Rana) are adjusting camera views to figure out where the spots are on the MC mirrors. The evidence suggests that when the mirror suspensions are adjusted so that the resonant beam is hitting the center of each optic, the transmitted beam is far from getting through the downstream faraday isolator, and vice versa. They are working on some optimal compromise, until we vent again to move the mirror between the MC and the FI accordingly.
  • Archiving of trend frames and /cvs/cds stopped when fb40m was rebooted; it was restarted.
  • we have ordered new spare hard disks for the DAQ RAID array.
  • Our RGA works, and the IP connection allows us to read out the AMU spectrum. However, the CRT display (which is also used to control the RGA settings) broke a few months ago. Rob has been figuring out how to change the settings using the IP connection, and is making progress. Bob has a spare Dycor CRT and will see if it works.
• IFO modeling:
• Lab Infrastructure, Bake lab:
  • Steve will be on vacation this week and next, so Bob will be looking after the care and feeding of the 40m.
  • Bob is doing some Lab clean up for Monday's visit by the First Minister of Scotland.

October 5, 2006

• DC detection and vacuum squeezing development:
  • Rob and Royal found several problems with in-vac cabling, fixed by Bob.
  • Rob, Sam and Valera completed the installation of the OMC optical train, including alignment of all optics, testing of OMC locking, DC photodetectors, electronics and picomotors, and re-establishment (roughly) of the PSL launch beam into the mode cleaner. The squeezer mirror picomotor is a bit cranky and can take hours to move into or out of position.
  • Osamu, Steve, Rana and Valera tweaked up the OSEMs on MC1, MC2, MC3, BS, PRM, and lastly, SRM. All OSEM signals are centered in their range. Improvements in POS/pitch/side couplings ranged from noticable to negligible. All EQ stops adjusted so that the spacing between stop tip and optic is ~2 mm for each.
  • Steve and Rob replaced the flange on the north door of the OOC with a window to allow the OMCT and OMCR beams to exit to the AS table for monitoring.
  • Today (Thursday), Steve pumped down the interferometer, with no problems. By 10pm, Rob, Sam, and Valera had the input mode cleaner locked, and were working on getting the transmitted beam through the faraday isolator (the OSEM work on the mode cleaner suspensions meant that the alignment to the Faraday was lost).
  • Ben found out the cause of last weeks issue with the 40m in-vac wiring. It all stems from the fact that the flange feedthrough is male on both sides, and a pin-flip happens when you pass into the vacuum. It turns out that the older in-vac wiring (flat ribbon cable, no shield) was wired up with the pairing starting at 1-14, 2-15 ..., with pin 13 by itself. This means that in the old cabling, modules in air need to be wired up with pairs starting at 13-25, 12-24... with pin 1 by itself. The new (twisted pair, overall shielded Cooner wire) cable fixed this anomaly, and its pairs start at 13-25, 12-24... The shield is tied to pin 1 in the vacuum, which is pin 13 in air. This special in-vac cable pairing means that each signal is always twisted with the correct partner both in air, and in vacuum. The problem arose when the 40m got the new style of cable, yet kept the old pin pairing. This was done because we have many LIGO I controls (satellite amps, etc.) that needed the old pinout. It just means that LASTI and the 40m have different cable pinouts, and if you design a module for one installation, it's not compatible for the other. You'd end up transmitting a signal on the shield, and the twisted pairing would be shifted down by one.
  • Ben has begun testing the QPD Whitening Interface board. So far all of the channels work, and seem to have the correct transfer function.
  • Ben designed and sent off the front panel for the QPD Whitening chassis. It should be here soon.
  • Bob is helping Rich with installation and wiring/cabling of the DC readout control electronics.
• IFO commissioning, Electronics, controls, computers:
• IFO modeling:
• Lab Infrastructure, Bake lab:
  • Bob has started making OSEMs for Janeen. She has ordered 8 dirty OSEMs for the UK to do testing and 8 clean for LASTI.
  • Bob finished the bake job for Betsy and has shipped the PMC parts back to her.

September 28, 2006

• DC detection and vacuum squeezing development:
  • We are vented, Steve and the crew have removed access connectors and chamber doors. Rana assembled a nice photo-essay in the ilog.
  • Rob has installed the DC readout beamline in the output optic chamber. Go and Osamu put in the translation-staged pickoff mirror for the squeezed vacuum beam.
  • The in-vac cabling was found to have one shorted pin, so that the OMC PZT could not be controlled. Bob, Dan, Ben and Rob worked to get it fixed / patched around, and now the OMC locks. Ben and Rich have updated the schematics / pinouts of the in-vac wiring to reflect this.
  • Rob and Sam aligned the beamline using back-propagating fiber-fed laser light from the DC PD all the way through the beamline back to the sinal mirror, then retroreflected all the way back to the DC PD. They also worked on leveling the optical table.
  • Steve noted that the kapton used to insulate the DC PD has non-vac-compatible adhesive on it. So this has been pulled out of the vacuum chamber, the kapton and adhesive removed, replaced with new kapton with no adhesive, and it is being prepared for re-installation into the vacuum chamber and re-establishment of the alignment. Rob set up a jig to precisely recover the previously aligned position of the DC PD mount, to make the process easier.
  • Riccardo is stuffing and testing his photodetector PCBs, with help from Ben.
  • Riccardo gave a nice talk on his balanced photodiode homodyne detector development for the squeezing project. He's finishing up his final paper and preparing to return to Italy after a job well done.
• IFO commissioning, Electronics, controls, computers:
  • While we are vented, Osamu, Steve and Dan are working on adjustments to the OSEM positioning on the suspensions for MC2, BS, PRM, MC1, and MC3. They are reducing the overall noise, positioning the OSEMs to the middle of their range, and reducing POS/SIDE coupling. Osamu and Rana are checking the results using excitations and free-swinging spectra.
  • The OSEM crew has discovered that the "earthquake stop" screws on our suspended optics are a mix of viton tips, springs, and ... nothing! They are busy making viton tips for this last category. They are adjusting all earthquake stops on these 3" optics to be 2mm away: not so close as to charge up the optic, not so far as to be ineffective in case of earthquakes or other control problems.
• IFO modeling:
• Lab Infrastructure, Bake lab:
  • Bob has started Betsy's bake job on the PMC assembly parts.
  • Bob is cleaning 6 stainless disks for an epoxy test, requested by Helena.

September 21, 2006

• DC detection and vacuum squeezing development:
  • The broadband noise on the DC PD was mis-estimated; it's actually lower than we thought, and is acceptable. There is still a lot of spikes in the noise, presumably associated with grounding problems, and the LP filter added to the preamp doesn't help much. But it's good enough for now. The preamp has been returned to the vacuum nipple, refilled with krypton, and the nipple sealed up.
  • Rob installed a laser and some optics on the AS table in preparation for aligning the DC readout optical train through to the SRM.
  • Steve and Dan began the vent on Tuesday at 11, and by late afternoon they started opening doors and removing the access connector between the BS and ITMY chambers.
  • The pumpdown caused the oplev signals to shift, and they are being carefully watched.
  • Rolf and Alex have begun the installation of the DC readout controls. More work will continue next week. They were able to bring the controls computer on line to test with the Framebuilder. The Framebuilder can now accept data from both the reflected memory network and the Myrinet network.
  • Bob is building cables for the DC readout controls, and temportary cables for the in-vac alignment work.
  • Go realigned the SHG beams and increased the green light power by a factor of 2. He then realigned the pump beam into the OPO and lowered the crystal temperature, getting the parametric gain up to 30.
  • Go measured the beam profile of the seed output and LO at the homodyne photodetectors in order to improve the homodyne visibility.
  • Riccardo is stuffing his balanced PD PCB boards and beginning tests. He's also frantically writing his paper and a talk for next week. He looks a little tired!
• IFO commissioning, Electronics, controls, computers:
  • Steve and Osamu opened up the MC2 chamber to improve the OSEM positioning on the suspension. Steve set up the oplev with MC2 exposed, do use as a reference diagnostic. Osamu spent some time fiddling with the MC2 OSEMs. The UR OSEM had a lower voltage than the others. They swapped the satellite amplifier with a spare, then the UR sensitivity became the same as the other OSEMs. The UR was also noiser than others. They swapped some cables and the noise got better. We need new, well-made 4-pin LEMO cables. Some of the in-vac OSEM pigtail cables were potentially close to the beam, so the OSEMs were rotated and the cables tied down. They adjusted all the OSEMs to give readings in the middle of their range (~0.8 V). Finally, they tried to decouple the back OSEMs from bounce and roll modes, and also decouple the side from POS motion; they didn't see much improvement, but it's good enough for now. ON to the other suspensions!
  • Dan replaced all the opamps on the LSC PD Interface Board in the MC rack with less noisy ones, to get better signals for the MC REFLPD and the MC RFAM PDs.
  • Dan replaced the ETMY Oplev laser diode with a small HeNe laser, and Rana measured the uncalibrated noise spectrum.
  • Alex put in a new martian network switch to handle the new controls computers.
• IFO modeling:
• Lab Infrastructure, Bake lab:

September 14, 2006

• IFO commissioning, Electronics, controls, computers:
  • Last week, Rana again noticed that the RFAM coming out of the Mach-Zender was very large, and very sensitive to poking with his finger. He found the EOM mounts to be less than stury, and thermally not stable. He tightened screws and did a few more tweaks. Osamu checked the alignment of all the EOMs and the input beam to the MC. We should plan on replacing the EOM mounts, re-laying out the MZ, and re-aligning the whole IO path, sometime after the vent.
  • Osamu and Rana will do a final round of IFO locking, taking baseline readings of all power levels and beam positions, in preparation for our upcoming vent.
  • Steve set up the MC2 oplev QPD to instead monitor the MC2 main laser transmitted light. However, we changed our mind, judging the oplev to be a more useful monitor during the vent. Rana will re-setup the MC2 oplev. Then we'll switch back to transmitted light, after the vent, since the MC2 oplev has never been very useful.
  • Steve noted that the beam path to the IOO POS and ANG QPDs was clipping, so he re-routed the path around the persicope. He then re-established good signals on all the reference QPDs (IOO POS and ANG, IP POS and ANG, all oplevs).
  • The lenses that mode match the PSL beam into the MC were not AR coated. Osamu and Steve replaced them with new AR coated lenses, and re-established the beam pointing into the MC. The visibility was measured to be 92% (somewhat lower than Osamu remembered from years ago). However, we now have 14% more light into the MC.
  • Rana plans to measure the offset of the beam spot on the ITMs. If ITMX is offset in x (horizontal, transverse to beam), we will move it a bit. Dan moved the ITMX horizontally by 6 mm last summer, and the beam spot as seen on the CCD monitor appeared better centered. This time, Rana is measuring more accurately by dithering the ITMs in pitch and yaw with the arms locked, and observing any angle-to-length coupling.
  • Osamu is writing up a detailed procedure for optimizing the OSEM positions on the suspensions, in preparation for adjustments to the OSEMs on the MC2, BS, PRM, SRM suspensions during the upcoming vent.
  • Rana has proposed that we replace our cheap, noisy oplev diodes with HeNe lasers. Dan measured the RIN of the HeNe laser, getting ~ 1e-6 V/rtHz baseline with some large spikes in the 100-500 Hz region.
• IFO modeling:
• DC detection and vacuum squeezing development:
  • Bob has prepared all the in-vac cabling for the DC readout beamline.
  • Osamu checked that the DC response of the two TT PZT mirrors behaved as expected; all the axes moved as expected (pitch for pitch and yaw for yaw) with ~ few mrad/100V.
  • Rob has completed the assembly of the in-vac beamline, on the flow-bench. He's wired everything up and done lowest-order testing. He has dither-locked the OMC using analog electronics, using both forward-propagating laser light and backward-propagating (from the DC PD back towards the main IFO). Handles have been installed on the breadboard so that it can be lifted up and placed in the vacuum chamber. Rob has written down a fairly detailed installation procedure in the 40m Wiki. Aside from the DC PD problem (below), he's ready for installation.
  • Ben, Bob and Rob assembled the DC PDs and in-vac resistors, and wired it all up. Rob discovered that one of the DC PDs was broke (no signal). It turned out that one of the PDs had been damaged either in the de-canning, or the cleaning phase. Ben brought over replacements, de-canned them, they were cleaned by Bob and installed by Rob, who verified signals.
  • However, Rob and Rana measured the noise and found it to be very large (~ 5x Johnson noise), with lots of HF spikes. Rana and Ben are working on mods to the in-vac amplifier electronics to filter out the spikes. Because of this, the vent was delayed from this Thursday to next Monday.
  • Rob and Rana are compiling a list of "lessons learned" from the DC readout assembly (even before we commission it!) on the 40m Wiki. Alan is assembling lots of related info on the DC readout experiment, on the Wiki.
  • Rolf has begun installation of the DC readout digital controls in rack 1X5: the Sun computer, IO chassis containing ADCs and DACs, AA chassis, master and slave timing modules. We hope to get in enough equipment this week to start running the I/O and software. Once the computer is up, we need to start testing the new Framebuilder code to verify it can accept data from both the existing reflected memory network and the new Myrinet network.
  • Bob, Ben, Rolf, and Riccardo are working on the cabling, ethernet, and fiber optic runs required for the DC readout digital controls in rack 1X5.
  • Rich has made significant progress on the Piezo Driver unit for the DC readout experiment. He anticipates being mostly finished with the PCB layout by tomorrow. The bill of materials is complete, and all the parts can be obtained on short turn around (overnight) from Digikey.
  • Bob has been ordering and collecting the supplies that we need for venting.
  • Go used the video output of the existing spectrum analyzer HP8591E to noise-lock the squeeze angle. About 2.5dB - 3dB of squeezing was observed at 900kHz.
  • Go is working on realignment of his squeezer optics and optimizing the OPO temperature. He also replaced the PPKTP crystal with a new one coated by Tafelmaier. After some effort, he got back to a parametric gain of 3.
• Lab Infrastructure, Bake lab:
  • Although the RGA display did not survive the power glitch of Sept 1st, both the RGA and its logger did survive. Thus we continue to have RGA spectra logged to disk. We'll continue in this way until after the vent, then decide on whether to replace the RGA with the SRS one.
  • Steve mopped the IFO hall with Dursban ant poison in preparation for the vent.

September 7, 2006

• IFO commissioning, Electronics, controls, computers:
  • A campus-wide power glitch on Friday 2pm brought us down for the rest of the day. Alex came by to help bring up the martian network. We got some good experience doing a cold-start. And we learned about critical systems that should be on UPS backup, but weren't. Now they are!
  • Rana did a new sweep of the NPRO temperature and found that the SLOW temp was not at the right place; we were in a multimode region where the NPRO is oscillating. He moved the setting to a place where the NPRO power is maximized and multimode oscillations are absent.
  • Rana realigned the PSL beam into the MC, and also realigned the MC refl and trans paths. He turned the boost back on the MC up script, and did some more optimization.
  • Ran began working on lock acquisition again, but discovered that the testpoints were not working so none of the lock acquisition scripts could be exercised. Alex came by on Tuesday and fixed it.
  • Rana and Sam checked out all the length sensing RFPDs. The double-demod PDs (PD7 and PD8) were not working; they found that the RF amps weren't plugged in, and also found a bad cable on PD7. Work is still in progress.
• IFO modeling:
  • Monica is returning to Orsay, where she will work more closely with her AdLirgo colleagues on modeling. Bon voyage!
• DC detection and vacuum squeezing development:
  • The focus this week is last preparations for vent to install DC readout optics. We plan to vent on Monday, and begin work on Tuesday.
  • Steve will check the initial pointing of the PSL beam, and establish baseline readings on all the QPDs, on Friday.
  • Osamu and Rana will work on adjusting the OSEMs on the noisiest suspensions: MC2, BS, SRM, PRM, and maybe also MC1 and MC3.
  • Rob and Sam are progressing on pre-alignment of the clean DC readout beamline on the flow bench. They are assembling the controls that they will use to lock the OMC and bias the TT mirrors. They are readying the fiber-fed laser for alignment. Rich has the HV PSs for the PZTs. They are working towards being ready to install into the vacuum chamber by mid-week next week.
  • Ben ordered and received a wide assortment of resistor values for the in-vac DCPD electronics. Rob chose two 250 ohm resistors; Bob cleaned them and wired them to the DC PD mount. He wired these to the in-vac pre-amp nipple, and from there to the 25-pin mounting bracket.
  • Rich received the repaired tip-tilt PZT from PiezoJena for the 2nd TT steering mirror. Bob cleaned it and is making the in-vac wiring. Rob is preparing the two TT mirrors in their mounts.
  • Bob has built all the required in-vac wiring: for the DC PD, the OMC PZT, the two tip-tilt PZTs, the OMMT picomotor, and the squeezed vacuum mirror picomotor.
  • Ben has the new rev of the QPD Whitening board all stuffed, has developed a test procedure, and will begin testing shortly. This board will go in one of the new style chassis which will be ordered soon.
  • Rich and Riccardo are preparing to assemble the new DC readout control electronics and cabling at the 40m. We hope to have the computer, IO chassis (including ADCs and DACs), AA and AI chasses, and timing master & slave in place by next week. We hope to also have a RFM card and connection to our DAQ system. (That leaves the QPD whitening board and PZT driver boards, which will come later).
• Lab Infrastructure, Bake lab:
  • In preparation for venting, Steve has done some work on the vacuum system, installing a larger drypump to have more maglev backing pump power. He's also tested the cranes. He will mop up with ant poison at the end of the day on Friday.
  • Dan is back from Burning Man. Rana asked him to move the 40m Wiki to make it separate from the AdLIGO wiki.

September 1, 2006

• IFO commissioning, Electronics, controls, computers:
  • The EPICS database for the new MC board had some problems; Ben fixed it, and Sam checked out that all controls and readbacks seem to work correctly.
• IFO modeling:
  • Monica presented the status of her work on e2e and at the 40m, in a meeting with her Virgo colleagues Patrice Hello and Fabien Cavalier, who were visiting earlier this week. Her talk is here. She is working on LIGO reports on her mirror velocity measurements and her work on the MC WFS Guoy telescope and servo diagonalization. She returns to France next week.
  • Osamu has been working with Kentaro on modeling of new RF control schemes for AdLIGO, validating with Optickle, Finesse, Corbitt-code. He also did a study comparing different arm cavity g-factors to reduce the severity of alignment instabilities.
• DC detection and vacuum squeezing development:
  • We planned to vent for installation next Tuesday, but we're not ready with the 2nd TT PZT. We got a spare one from LLO, but they want it back. We await delivery of two TT PZTs from Piezo Jena. Word is, we'll get one (the repaired one) sometime next week; so we'll postpone venting until that is received and checked out.
  • Rob and Sam are assembling and aligning the in-vac optical beamline on the clean bench. the only hitch was a missing mirror clip. Ben made one, and Bob cleaned it (twice). They're now starting on the OMC assembly, electronics, alignment.
  • Status of controls electronics for DC readout: Ben has completed an updated draft of the system wiring diagram. Computer, PCIX IO chassis, ADCs, DACs and REFMEM are all in hand and under test by Rolf & Alex. AA and AI boards are assembled and under test. QPD whitening board is stuffed and Ben will test it soon; he'll use a new LIGO Rack Mount Chassis which will be ordered soon. DC satellite box and in-vac DCPD amp are in hand and in use. Rich is still working on the PZT driver boards.
  • Ben changed to different opamps on the DCPD amplifier board to get lower overall noise, did a functional test, and updated the schematic.
  • Ben and Steve installed the DCPD amplifier board into the vacuum nipple, filled it with krypton, and sealed it up. Rob will test it on the clean bench in preparation for installation.
  • The new rev of the QPD Whitening board is stuffed now, and ready for testing. Ben will write a test procedure, then it can be tested in the back shop.
  • Ben cut the cans off two GAP2000 PDs that we can use in the vacuum DCPD.
  • Rich chose, and ordered, some Kepco high voltage DC power supplies for the PZT drivers for the OMC and TT mirrors.
  • Steve is preparing the lab for DC readout installation. He has cleaned out two small electronics racks and cleared out the cabling; the racks are ready.
  • Riccardo continues working on balanced PDs for the squeezing experiment. He assembled his circuit on a breadboard and set up an optical system. He then measured the relative QE of 12 PDs, (ranging from 82-87%, in good agreement with expectations), and estimated the errors. He also measured the angular dependence of the result, and found it to be roughly flat out to 60degrees. His PCB design has gone an iteration with Chris Wipf, and is ready to be sent out to PCB express for fab. Ben will help with that.
  • Go and Shally designed and built a spectrum analyzer for the squeezing experiment. It is now completed and tested, now being integrated into system, with help from Riccardo.
  • Go is now realigning the squeezer optics to reestablish squeezing with the new spectrum analyzer.
  • Go is working on two papers describing the squeezing experiment.
• Lab Infrastructure, Bake lab:
  • Steve checked that particle counts in the back clean room (where the DC readout beamline is being assembled) are near zero.
  • Steve plans to add more fore-pump power behind the main TP1 maglev turbopump to avoid overheating during pumpdown. He received some vacuum hoses to make these mods.

August 25, 2006

• Last week was the last for 40m SURF students Jenne Driggers, Darcy Barron, David Malling and Royal Reinecke. They all did terrific jobs and will have impressive final papers. It was a pleasure having them all and we wish them the best. Also, this is Shally's last week before returning to Rochester, but he'll come back when the squeezing experiments start.
• Rob presented the status of work at the 40m at the LSC meeting.
• IFO commissioning, Electronics, controls, computers:
  • Osamu and David measured the noise spectrum in the full configuration except for carm offset, low power, with common mode servo fast path on. The noise floor remains the same, but many big peaks are gone; 60 Hz harmonics all still there, due to AP166. The unchanged noise floor points to electronics noise problems.
  • Jenne used her AM laser test setup to take swept-sine measurements of ALL RF PDs in the interferometer (!) (except for the IO WFS sensors). They all seem to behave as desired / designed.
  • The MC spot has been wobbling ("blowing in the wind") ever since the new MC servo board went in. Rob found several problems associated with the reboot that failed to restore proper settings for gain, input matrix, etc. After fixing several things, the "wind" got calmer, but still needs some attention.
  • Steve and Dan found a bad board channel reading out the IPPOS QPD-ICS Interface board. We need this, before vent, as a reference for initial pointing! They replaced the board with a spare and centered the spot. Ben looked at the board, saw evidence of a damaged op amp, and gave it to Todd for repair.
  • We had several problems with the front end network in the last couple of weeks, requiring Alex's help. Last week, the RFM card in the IOWFS control computer (c1iovme) failed, bringing down the entire RFM network. Alex borrowed a card from the ASC system, and recovered the system. Several days later, he brought a replacement RFM card for the ASC system, and he and Rob brought that system back up, so that we could once again control the input steering mirrors and align the interferometer.
  • At Sam's behest, Ben looked at the database for the new MC Servo, and found that the database didn't match the wiring by one channel. This shifted all of the 3113 channels down by one, and made no signal flow to the third channel. Ben fixed the database, and Rob modified the saverestore file and rebooted the processor. Seems to be ok.
  • Monica and Dan are implementing a WFS matrix diagnostic and ran into problems. They consulted Matt Evans about this, who gave them a Perl script which can form a basis to work around the problems. Monica is writing a technical note about WFS diagonalization.
  • Rana implemented a new modified elliptic low pass filter called ELP35 for the ETMX oplev servo, which reduces the self-inflicted noise considerably above 40 Hz. Royal implemented this new filter in all the other TM oplev servos, and measured/calculated calibrated noise spectra for all test masses, before / after the change.
  • Sam measured the MC length using two different methods and got more-or-less consistent results: 13.547003(4) m. He set the RF frequency to 33,194,700 Hz for maximum transmission.
  • Rob added the new MC servo board channels to the conlog. While he was at it, he ran update_chanlist (hadn't been run in quite a while) and added 920 channels to conlog.
  • Rana, Rob and Sam worked on DRMI locking and made progress, but were hampered by lots of RFAM on the double-demod signals. The source appeared to be from the Mach Zehnder, which they tweaked to minimize the 29.5, 33, & 166 MHz RFAM peaks. They brought the system into lock with CARM offset, handed off to DD signals and CM servo, but had trouble with the CM AO path. Work in progress.
  • Rana hooked up a frequency counter with a GPS-based external reference, setting up a BASOLUTE frequency standard. He used it to check the absolute value and stability of our Marconi RF signal generators.
  • Steve reports that the laser chiller still appears to manufacture ~10cc of water per day; Steve has to drain it periodically. A complete mystery!
• IFO modeling:
  • Rana used Optickle to study the expected DRMI signals. He made PRC/MICH/SRC sweeps, and got some insights into what signals work best for each loop.
  • Patrice Hello and Fabien Cavalier from LAL Orsay / Virgo will be visiting to talk about collaborative work on AdLIGO / AdVirgo optical configuration & control. We'll meet on Monday 2-4pm in the SCR; all interested parties are welcome to attend.
• DC detection and vacuum squeezing development:
  • Rob (with Darcy, before she left) is re-assembling the in-vac DC readout beamline on the clean bench, and preparing for pre-alignment. Hope to be done and ready for installation in a week. We will vent on September 5.
  • Rich presented his design of the TT and OMC PZT driver board. Getting the go-ahead, he has started the layout of the unit.
  • We received a spare TT PZT stack from LLO, and Sam is talking with Ken to make sure we know how to wire it up. We have two more TT PZT stacks coming from Piezo Jena (a repaired one and a spare), so that we will be able to return the borrowed PZT to LLO in good time.
  • All the other CDS hardware and software are almost all ready. Anti-aliasing chassis is fully assembled and under test. Anti-image chassis board has been assembled; it needs to be put into a chassis and tested. IO chassis rebuild is mostly done. QPD Whitening chassis, latest version of the board in; figure at least the rest of the week for assembly. DC readout Satellite box is in hand and in use. ADC and DAC interface boards are all assembled in the IO Chassis. Jay and Rolf have tested the controls software and screens, and the RFM link to our existing framebuilder. Anticipate being finished with work in the back shop next week. Computer and AA/AI chassis for OMC should be ready to install in the next few days.
  • Ben is drawing a schematic for Bob so he can make the in-vacuum wiring for the DCPD from the vac. nipple to the PDs. Bob has prepared and vac-prepped almost all the remaining in-vac cabling.
  • Go and Shally tweaked the PPKTP OPO temperature and got a parametric gain as high as 21.4. They got up to ~ 2dB of vacuum squeezing at 800 kHz.
  • Go was at the SPIE conference last week. He and Shally are now designing a single frequency (800 kHz) power monitor to be used for noise locking.
  • Go and Shally tested the squeezed vacuum mirror picomotor translation stage. It works fine but is slow: it takes ~ 45 minutes to move 1.5". Bob is now preparing it for vacuum, including the in-vac wiring.
  • Riccardo Ciolfi has his schematics ready for a balanced PD for the squeezed vacuum experiment, and is finalizing the PCB layout design, with help from Ben. All components are in hand. He is testing single PD circuit on a breadboard.
• Lab Infrastructure, Bake lab:
  • Dan continues to add the content of the racks to the wiki. This is just a preliminary survey of the components in each rack, with the intent of leaving a framework that can be flushed out with more information later.

August 3, 2006

• All the SURF students have prepared 5 minute talks for their trip to LHO and are off on their junket.
• IFO commissioning, Electronics, controls, computers:
  • Rana and Sam are going through the LSC RFPDs and re-phasing them (since the cables have all been modified to strain-relief them). Several single demod RFPDs have been rephased. They are encountering difficulty in phasing the double demod signals. Work in progress.
  • Steve is measuring the response of the electronics chain from RFPDs to mixer to whitening boards to ADC, by injecting signals at different frequencies. He's getting the response he expects, dominated by whitening board TF.
  • Royal measured the oplev P/Y dark (electronics) noise. She got generic whitening boards from Wilson House and will modify and implement them for all 4 oplevs, to try to reduce the electronics noise.
  • Royal measured the laser intensity noise of all 4 oplev lasers. She also measured the TF from input current to output power. Power could be modulated up to high frequencies, > 100 kHz. But intensity noise is not quite a problem right now.
  • Royal finished measuring the mirror motion due to servo for all 8 DOFs. Made a noise budget for all of these. And then through to DARM at the DC PD.
  • David, Jenne and Sam have implemented the new MC servo board; this required rewiring and debugging of the cross connect wiring, testing of digital controls with tester boards, testing of the board, etc. All seems to work ok, but there are problems with readback of some MC signals (err point, fast channel, and MCL monitors) into the EPICS screen.
  • The MC is now locked and working fine. After some PD phasing and tuning, they measured the MC loop gain: the UGF is up to 80 kHz. It could be better: the MC OL gain has odd resonances at ~80, 90 kHz, and seems to fall off at ~230 kHz, probably due to PMC. Darcy will measure the PMC finesse. One day we'll replace the PMC with a low finesse one. Sam says that more gain could be obtained by further tweaks to the AOM, but this is difficult to do because of all the stuff in the way on the PSL table.
  • The MZ stopped locking. Sam, with help from Jenne, David, and Rob, debugged the system and found a bat XYCOM cable connector. They gave it a good squeeze, and the MZ is now behaving.
  • David constructed some insulating houses for the OSAs on the AS ISC table. The temperatures on the OSAs are being monitored and logged. Steve will also move the particle counter to that table; it has a coarse temperature readback. David will watch all these, and the OSA display, to see if the temperature fluctuations are real, are reduced by the insulation, and/or are correlated with the OSA peak drifts.
  • Jenne's AM laser is all put together with flex tubing. She used it to check the MZ RFPD. She will measure all the RFPDs when she gets back from LHO.
  • Dan and Monica rotated the MC WFS's by 45degrees a few months ago, and has now changed the MEDM screens to correctly represent this. He needs to also change the MCLock screen, DC signals.
• IFO modeling:
  • Jenne has completed a working Simulink model of the radiation pressure-induced angular instability in the mode cleaner, and has checked her results against the predictions from Sigg's paper. She will generate OLTFs with different input powers; the pitch/yaw pendulum response will be altered in the presence of the optical torques. She'll then try to measure these with the real MC, again at different input powers and different angular offsets of the mirrors.
  • Osamu is working with Kentaro on alternative length control schemes for AdLIGO. They are exploring different RF sideband configurations, and also ways to implement variable signal detuning. So far, 9 and 45 MHz works with fixed signal detuning, but implementing variable signal detuning is more difficult. He is preparing a talk for LSC meeting.
  • Seiji learned the current lock acquisition procedure from Rob, and is documenting it and studying it in simulation (Finesse) to see how our LA procedue works and how to improve it.
• DC detection and vacuum squeezing development:
  • Darcy and Rob disassembled the DC readout beamline and gave all the parts to Bob. Now they're preparing for re-assembly and realignment on the clean room flow bench. They are running the NPRO laser through fiber to get it to the clean room.
  • Bob has been preparing detailed travellers for all the DC readout parts. All parts cleaned, OMMT in for bake; will put all the OMC and other parts in by the end of the day. Baking copper at 120; anneals at 145. Working on in-vac cables and connectors.
  • No report from the Wilson House group; they're at a meeting at MIT.
  • Shally and Go made lots of progress on the vacuum squeezer. They Rebuilt the OPO oven and cavity. They still saw elliptical modes, but then they turned the (ostensibly spherical) output coupler, and got a circular mode. They temperature-tuned the OPO by maximizing green output. More alignment and temperature tuning got parameteric gain up to 7.8. They set up the homodyne detector with fiber-fed LO and seed field. Fringe visibility of 88%; good, to be improved. They looked for squeezing by dithering the seed beam path length, and antisqueezing observed. Hopefully can see squeezing soon. With 3dB of squeezing, they will do quantum noise locking to lock the squeeze angle and the OPO.
  • Latest news: they now observe 1 dB of vacuum squeezing (and anti-squeezing). This is the first observation of CW squeezing at 1064nm with PPKTP in the world. The seed was completely blocked. So, this is the obervation of pure squeezed vacuum.
• Lab Infrastructure, Bake lab:
  • Dan continues to flesh out the new 40m Wiki. He encourages anyone who pulls out a board to take a photo of it for the Wiki.

July 28, 2006

• IFO commissioning:
  • David and Osamu continue to develop the noise budget procedure and code. They are studying the effect of whitening on the noise spectrum.
  • Royal measured the oplev laser intensity fluctuations and determined that their contribution to the oplev pitch/yaw noise, on the order of 1e-12 rad, is negligible compared to other noise sources (jitter, electronics). She then quantified the jitter noise on mirrors due to the imperfect filtering of the oplev servos at high frequencies: ~1e-9 rad/rtHz at 100 Hz. She's working on the electronics (dark) noise, and is continuing her study of the mirror motion and noise budget in the oplev system.
  • Rana and Royal are going through the oplev whitening electronics and finding differences, and undesirable whitening stages. They designed some digital compensation filters to patch the problem, which seems to work. Rana replaced some filters on the ETMX oplev board, changed the digital filtering, and reduced the jitter noise above 100 Hz by 100x. More work is needed on the digital filtering to get a better loop design. We should replace the existing boards with RevB oplev interface boards on all four test mass oplevs.
• IFO modeling:
  • Jenne continues to work with Rana on a simulink angular instability model for the mode cleaner.
  • Osamu continues to work on e2e simulation of the AdLIGO alignment instability.
• DC detection and vacuum squeezing development:
  • Darcy and Rob completed their measurements of the DC readout in-vac beamline, including measurements of the OMC finesse as a function of temperature. Darcy is writing it all up, complete with photos, plots, etc.
  • Darcy and Rob have completely dismantled the DC readout in-vac beamline (TT mirrors, OMMT, OMC, DCPD, breadboard) and inventoried all the parts. Everything is now ready for Bob to clean and bake.
  • Rob is fleshing out a detailed schedule / plan for venting and DC readout installation.
  • Go's in-vac components (picomotor translation stage and two mirrors/mounts) are now in Bob's hands and are being cleaned/baked.
  • Ben's new QPD Whitening board is being stuffed in the back shop currently. Jay's AA and AI interface boards are also in the queue.
  • Ben has finished all of the machining on the DCPD mount, including mounts for the black glass beam dump and the power resistor. The whole assembly is ready for cleaning in its current state. Because the mounts are going into the vacuum, and won't be touchable, the whole assembly is now in the machine shop being dimensioned including all of the modifications. This way, if anywhere else wants a DCPD, Ben can have ones made in the future with one phone call.
  • Ben has almost finished the system drawings for the OMC installation. He has finished the drawings from the point of the AA and AI chassis through to their modules in vacuum. The drawing at this stage was given to Bob so that he could order the in-vacuum cables from Accuglass.
  • Rich, Sam & Rob decided we don't need the fast shutter to protect the DC PDs, and it complicates the design.
  • From Rolf:
    • The OMC software is up and running. I still want Rana to come over and look at it to verify that it is correct. Major items added during the last week to overcome a few hurdles:
    • M. Evans was in town and gave me some software for the fast generation of sine/cosine signals (5 sine/cosine signals required for OMC). I built a simulink part and then added the function to the code which takes the matlab output file and produces CDS front end code. It looks like these new oscillator signals run very well and all are produced in well under a total time of 1usec per 32kHz control cycle. The code, as written into the front end controller, allows for frequencies and amplitudes to be changed on the fly. So, we will not need to tie up 10 AWG channels to produce these signals. This could possibly also be applied at the sites to generate the cal lines or at the 40m for LSC dither locking signals?
    • The OMC controller to be installed at the 40m lab is based on our new prototype designs that use Myrinet as the realtime network instead of reflective memory to communicate to the Framebuilder. To allow data to be merged from the old and new systems, Alex modified the Framebuilder software such that it can take data simultaneously from both types of network. This will require further testing once the OMC system is installed.
    • The new OMC controls still need to use reflected memory to communicate signals to the existing LSC controller. Driver code was written and added to the new system to use reflective memory for this purpose and this seems to work fine.
    • Rich finished prototyping the HV driver for OMC length and tip-tilt drive to be used on the DC readout. A production version is now being created. Parts must be carefully researched to ensure availability and applicability. This is ongoing. Thanks to Paul Schwinberg for his suggestions on the design.
    • Rich is contacting the outside testing service I found during the LASTI build to see if they can help out on testing some of the OMC stuff.
  • Go & Shally put the green pump beam into the OPO and got some parametric gain, ~ 1.2. Aim for 10 or more. They need to tune up alignment and mode matching, using correct OPO cavity length. they suspected a probem in OPO cavity, disassembled it, noticed some heater burn, and repaired it.
• PSL:
  • The PSL FSS slow loop got into an infinite loop, requiring a reboot on the PSL vme cpu. Peter King came to look at it but doesn't know why / how it got into that state. Rana figured out how to reset it without having to reboot the cpu.
• Electronics, controls, computers:
  • Ben gave Sam the cross-connect drawing and installation instructions that he made for the new Mode Cleaner servo. Sam will work with Bob in installing the system, and will change the screens and code to match the new board.
  • Ben helped Sam make some tester boards for the new Mode Cleaner servo installation. With these boards, the whole system can be checked in short order.
  • Sam, Dave, Jenne and Bob worked on the implementation of the new mode cleaner servo board. They redid the cross-connect wiring for the new board and for the LSC photodiode interface board. They modified the epics databases ioo.db and mc_lock40m.db. David checked all the sliders and buttons on the new MC Servo EPICS screen, and they all check out at the cross-connect. They labelled all the wires and cables, and are preparing to install the board and commission it.
  • David and Osamu measured the transfer functions of the dewhitening filters for the four coil drivers in each of the four test mass suspension controllers. They used the spectrum analyzer tools to fit to a zpk model, copied it into Foton and compared with the raw data. They then designed compensation filters and installed them into the digital filter bank, and tested that they did the job. The DARM signal is unchanged between digital and analog dewhitening filters.
  • Ben showed David how to balance the dewhitening filters offsets, so that switching dewhitening on and off happens smoothly. Ben did ETMY, and David did the rest.
  • David implemented temperature sensors in the optical spectrum analyzers at the AS and SP ports. The data are going into frames and trending. The femp drifts seem high (2-3 degrees C); they will check the sensor calibration. They plan to observe the OSA drifts and see if they correlate with the temperature drifts; if so, they will install some thermal insulation, and if necessary, heaters and servos.
  • Jenne is working with Steve to acquire more optics for her AM laser test system.
  • Steve and Bob completed most of the RF cabling upgrade at the AS table, the LSC rack, and the RF distribution rack. All cables are strain relieved, and Steve tested signals through all the cables.
  • As part of the RF cable upgrade, the Mach Zehnder cable was changed. Sam and Jenne re-phased the signal, and made a new cable of optimal length. The pk-pk error signal at the MZ servo board is now 273 mV, and the servo works real good.
  • Dan set up a new GC hub and ran ethernet cables to two connection-less WinXP computers in the office area.
  • Dan is moving our old, rather primitive 40m Wiki into the new 40m section of the AdLIGO Wiki and adding lots of content. Work very much in progress.
• Lab Infrastructure, Bake lab:
  • The air conditioning in the lab started to fail because the campus chilled water supply is overloaded, and the temperature went up to 81degrees; the laser chiller couldn't handle it. Steve reset the thermostat and opened doors, and got the temperature back down. We need a real AC system in the lab!

July 21, 2006

• IFO commissioning:
  • David and Osamu have recalibrated the ETMs and ITMs (in nm/count) using the michelson for the ITMs and arm lock for the ETMs. They get numbers in good agreement with the ones measured by Osamu back in February.
  • David and Osamu are working on measuring the coil driver noise with and without dewhitening (for the noise budget), and the dewhitening board transfer function (to get the digital compensation right).
  • Monica repeated her measurement of the ratio of the ITM/ETM transfer functions, which is useful for analyzing the noise spectrum.
  • Monica continues her WFS diagnostic tool development.
• IFO modeling:
  • Jenne and Rana are developing a simulink model of the angular control / instability in the mode cleaner.
  • Royal measured the oplev laser intensity fluctuations to be ~10^-6. She's calculating how that translates into P/Y measurement fluctuations, thence to the resulting power jitter at the AP, and finally, the effect on the DARM signal.
  • Osamu continues his e2e simulation of the alignment instability in AdLIGO with the quad suspension. He found a problem with the PSD module in e2e, and Hiro is working on it.
  • Monica prepared a LIGO-TN on her measurement of the test mass velocity, and comparison with e2e predictions.
• DC detection and vacuum squeezing development:
  • Darcy measured the transmission through the OMC to be 60%, due to anomalous loss/clipping at the small curved PZT-mounted mirror. Darcy and Rob rotated the PZT mount to reduce the clipping, and now Darcy measures T = 95%, which is reasonable for a cavity with a finesse of 200. Darcy, Sam and Rob are now taking a complete set of measurements of the in-vac DC readout beamline in preparation for disassembly for clean/bake.
  • Bob is ready for clean/bake of the DC readout beamline in the next week.
  • One of our tip-tilt PZTs is broken and has been sent back to Piezo-Jena for repair. Rich is trying to get them to expedite the repair (since we wanted to install in the vacuum any day now), to no avail. So Sam is trying to contact Ken Franzen to see if he will ship us a spare, to tide us over.
  • Rich is preparing a quote for the Piezo-Jena electronics to drive our two in-vac tip-tilt steering mirrors, as a backup for the driver board he is designing. Rich is wondering what the output noise is on these drivers, so Rob will measure it on the system we have. Rich's quote includes a spare tip-tilt PZT, because these things are ultra-fragile.
  • Rana assembled a block diagram of the front end control software we'd like to see for the DC readout system, including OMC dither-locking and steering mirror dither locking. Rolf has prepared a simulink block diagram of the system. An outstanding question is how the DCPD signal will get into our main interferometer LSC system.
  • Rolf is designing digital demod software for dither locking, and is consulting with Matt about how he did it at Virgo.
  • Ben's new QPD Whitening board is in the back shop awaiting stuffing. So are Jay's AA and AI interface boards.
  • The DCPD that seemed a little noisy last week was tested further, and it now appears to have an input referred noise of 6nV/rtHz. Ben might be able to lower this a tiny bit more by changing the 1st OpAmp to an OP27. He's also looking at putting in a zener diode for protection of bias voltage.
  • Bob did an RGA scane of the ceramic resistors for the in-vac DCPDs. He was surprised at how well they passed the scan, and it is his opinion that they are compatible with our vacuum system.
  • Ben has a little more machining to do on the DCPD mount to make a place for the resistors to be mounted, and to lower the possibility of capacitive coupling of the head to the base. This should be done by tomorrow. He's also machining a in beam dump holder and getting some black glass for the beam dump.
  • Ben and Rich are working on the system drawings for the DC readout installation, including the rack drawing. It appears that almost all of the electronics can fit inside our small rack by the output optic chamber (dubbed 1X5); the big computer will sit on top, and the power supplies must be located elsewhere.
  • Ben is thinking about putting a whitening stage in the DCPD satellite box (as requested by Rana), and also putting in a front-panel spigot for the DCPD signals to get them into the existing LSC system (since it is not trivial to get the signals from the new PCIX-based control system into our old RFM network).
  • Sam and Rob are determining the invac cable lengths so that Bob can start making cables. The cables for the tip-tilt PZT mirror on the BS chamber will have to run across to the ITMX chamber; that 110" cable run might be made by Accuglass.
  • Alex installed a Myrinet network card in our framebuilder; that's how it will communicate with the new PCIX-based front end control computer for the DC readout.
  • Go and Shally have redesigned and built a new heater system for the SHG, and have addressed a mechanical instability. Then they aligned and mode-matched the pump beam into the SHG and did an oven temperature sweep (peak at 114 degC). The SHG is now locked, and much more stable. Intensity fluctuations now down from 5% to 1%. The green beam is aligned into the OPO and they are now working on that.
• PSL:
• Electronics, controls, computers:
  • Bob has redone the RF cabling from the LSC RFPDs on the AP table (AP and REFL) to the LSC rack, with better connectors and strain relief. Steve has beeped them out, measured the losses, labelled them. Afterwards we had signals everywhere and could lock the interferometer. Rana is setting the double demod phases and gains. Next: redo the cabling on the PO and SPOB RFPDs, and at the RF distribution rack.
  • Sam and Jenne spent some time working on the input mode cleaner servo board, which convinced Sam that we should chuck it and replace it with a new version with the latest revs. We have this in hand, and now Ben is looking at wiring it up. It looks like it will be a simple cross-connect wiring and database change. Ben will give Bob the cross connect wiring list so that the changes can be made on Monday.
  • David and Sam continue to work on measuring, and stabilizing the temperature fluctuations at the optical spectrum analyzers.
  • Jenne continues to work on her AM laser system (for testing LSC RFPD response). She is ordering more optics.
  • Our DAQ system went down on Tuesday, Alex came and found a flaky RFM card in c1iovme. He rebooted just-about everything and it came back up; but we're keeping an eye on that flaky RFM card.
• Lab Infrastructure, Bake lab:
  • Bob has done vacum prep for a huge load of Accuglass cables for LASTI, and SS conflats from MIT. He's also building OSEMs for Gingin.

July 14, 2006

• IFO commissioning:
  • Rana, Sam and the students continue to work on locking the IFO in the full DRFPMI configuration. Jenne found and fixed a busted cable on AS166 RFPD, and now the signal is increased by a factor 50. This thing has been busted since January; its repair will surely facilitate lock acquisition efforts! After that, they locked all 5 degrees of freedom easily and brought the arm power up to ~ half-full, before noise sources stymied further progress. Under investigation.
  • Monica has diagonalized the mode cleaner WFSs and the system now works well. She is developing diagnostic tools for the control matrix.
  • Royal has calibrated the oplevs on all four test masses using the arm power variations as the mirror alignment is changed. She is studying the oplev servo noise and working to identify the sources of high-frequency noise (oplev laser, electronics). At low frequencies, there are indications that some of the damping servos are not working right. To be investigated.
  • Using her calibration of the test mass oplevs, Royal has measured the beam angular jitter in the interferometer, and extrapolated through the signal mirror and output mode matching telescope to predict the jitter that the output mode cleaner will see, and resultant transmitted power fluctuations. Rana used this information to specify the requirements for the alignment control and the output mode cleaner servo.
  • David and Osamu continue to develop the noise budget model. David added the model of the DRFPMI optical response to replace the simple Initial-LIGO-like cavity pole model. They remeasured the XARM noise spectrum with the dewhitening filters turned on; the noise floor at high frequencies was ~ 10^-17 m/rtHz, but with lots of lines.
  • David will begin measuring the analog transfer functions from the DAC, through the dewhitening filters, to the coil drivers. Then he'll tune the digital compensating filters and make sure that the switches do the right things. This could significantly improve the noise spectrum.
• IFO modeling:
  • Monica continues to develop the AdVIRGO e2e package, starting with just one arm. It locks at low power, and she is learning how to deal with the radiation pressure as the power is turned up.
  • Osamu in continuing to work on modeling the alignment instability in AdLIGO, with e2e.
  • Jenne is beginning some calculations to estimate the effect of the Sidles-Sigg angular instability in the 40m mode cleaner, to see if we can measure it.
• DC detection and vacuum squeezing development:
  • Sam and Darcy are making final measurements on the DC readout beamline prior to disassembly for clean/bake. The OMC transmission is only 50%, which they traced to a ~0.5% clipping of the beam at the small (7 mm dia) curved PZT-mounted mirror. They are working on various methods to fix this, but it might require ungluing the mirror and replacing it with a spare. They have remeasured the OMC mirror transmissivities (input couplers are 1.54%), finesse (213), total round trip loss (5.8%!), and other parameters.
  • Darcy found that our first TT PZT stack developed a resonance at ~ 80 Hz, with a strong pitch-yaw coupling. Opening up the case, they found a loose glue joint. They sent the stack back to Piezo-Jena, and Rich is ordering a spare stack. This will surely delay our vacuum installation significantly. Fortunately, Darcy reports that the 2nd TT PZT stack is fine, no resonances below ~450 Hz.
  • Rich is taking the lead on assembling the controls for the DC readout. He and Ben are reviewing the wiring diagram and filling in the details. They walked through the 40m this morning to better visualize the placement of the electronics and cabling.
  • Rana, using data from Royal, gave Rich the specs for the tip-tilt PZT and OMC PZT drivers. Rich has a first cut of the driver board. As a backup, he is ordering the commercial drivers from Piezo-Jena.
  • Ben's QPD Whitening board is back from PCB Express. It looks fine, and parts are arriving today to put it together.
  • Ben has a little more machining to do on the DCPD mount to make a place for the resistors to be mounted, and to lower the possibility of capacitive coupling of the head to the base.
  • The DC PD seems a little noisy, so Ben has brought it back to the Wilson House for more testing.
  • Ben got more of the ceramic resistors that we plan on using for the in-vac photodiode, and gave a good sample to Bob for RGA testing. The larger sample should let us know whether they are really vacuum compatible.
  • Rolf has the computer and PCIX boards (ADC, DAC, RFM) for the controls, and he and Alex are working on assembling the system. Jay will provide the anti-aliasing, anti-image and IO chassis interface cards.
  • Rana and Sam have drawn up a block diagram of the controls software to be implemented by Rolf (OMC length control and tip-tilt alignment control servos using dithering). Rolf had no objections, but he notes that several new software functions need to be implemented.
  • Go installed the new PPKTP crystal for the OPO. He saw lots of second-harmonic-generated green light from it, indicating that it has a high chi2 nonlinearity. He replaced the old SHG with the swapped-out OPO, expecting it to be more stable; after mode-matching and alignment, he locked it stably for an hour, and saw green light at room temperature. He's preparing the in-vacuum parts (one fixed mirror and one movable mirror) for injecting the squeezed vacuum into the asymmetric port. He verified that these mirrors remain highly reflective even at incident angles far from 45degrees.
• PSL:
  • The PMC transmission was oscillating ~5% diurnally and drifting low by 5-15% over the last couple of days. Matt touched up the alignment of the input beam and brought it up ~15%.
• Electronics, controls, computers:
  • Jenne has built a ~4 MHz notch filter and low-pass filter for the mode cleaner servo, in an effort to reduce high frequency noise that Rana believes is due to the beat between the IFO control sideband at 33 MHz and the mode cleaner control sideband at 29 MHz. She and Rana will implement it soon.
  • Bob is ready to install new RF distribution panels at the RF rack and the LSC rack, in the near future. It will take a couple of days and be somewhat disruptive.
  • Jenne has completed a mount for an AM laser to be used for PD testing. She measured the RF response of all the LSC RF photodiodes on the sensing tables, and all seem to have reasonable response functions. She's fixing an unwanted notch in the AM laser setup at 4 MHz and will retest all the RF photodiodes again. She'll also compare her AM laser test setup with the one at Wilson House.
  • The RF lines from our optical spectrum analyzers at the AP and SP drift about. To see if this was due to thermal drifts, David installed thermal insulation (foam jackets) and temperature sensors on the OSAs and sent them in to the DAQ system. If necessary, he'll set up heaters and thermal stabilization servos. Work in progress.
  • Rana has identified a code bug in the filter modules which causes the front-end cpus to run much slower when the gain is set to zero. Setting it to 1e-6 fixes the problem. rolf has been alerted.
  • Vuk copied over IFOtest from LLO to the 40m, and got it working. It performs automated checks of the various gains, switches, filter modules etc.
• Lab Infrastructure:
  • Steve replaced the failed RGA with a spare, baked out by Bob, and is in the process of calibrating it. It seems to give sensible results.
  • Steve arranged for a "doggy door" installed between the IFO hall and the clean room in Bob's lab, to facilitate the running of cables (and small parts) between the two rooms. The overpressure in Bob's clean room keeps a steady flow of clean air into the IFO hall.
  • Steve reports a large spike in the particle counts on July 4th.
  • Sally Sharaf, visiting from Rochester to work with Go on vacuum squeezing, received lab safety training from Steve, and an entrance eye exam.

June 22, 2006

• IFO commissioning:
  • Rana is back to work locking the IFO in the full DRFPMI configuration, while Rob is away on vacation in Sweden. Jenne is learning the ropes and the scripts.
  • Monica is using the WFS diagonalization code developed by Rana and Dan, and is preparing to diagonalize the MC WFS system.
  • Royal has been working on testing Mike Smith's oplev telescope, but she's having difficulty establishing a good spot on the PD when the beam gets through the thing.
  • Royal is developing a scheme to calibrate our test mass oplevs by wiggling the mirrors when the arm cavity is locked, and relating the transmitted power to the mirror angle.
  • Royal is exploring oplev configurations which suppress or cancel laser beam (mount) angle jitter.
  • David, working with Osamu, is learning the noise budget program. He is adding the model of the DRFPMI optical response to replace the simple Initial-LIGO-like cavity pole model. He ran on old data; the program ran through, and he's looking at results now.
• IFO modeling:
  • Kentaro is working with Osamu and Peter Fritchel on new length sensing schemes for AdLIGO, and maybe 40m. The double-demod offset needs to be understood better. He's checking the dependence on the RF sideband imbalance.
  • Valera is visiting from LLO, and is working on AdLIGO angular controls, optimizing the sensing matrix. He's also learning the 40m controls with Rana.
  • Monica continues to develop the AdVIRGO e2e package, starting with just one arm. It locks, but only with radiation pressure turned off.
  • Osamu in continuing to work on modeling the alignment instability in AdLIGO, with e2e.
• DC detection and vacuum squeezing development:
  • Sam and Darcy are trying to get the DC readout beamline fully checked out, so that they can disassemble for clean/bake in less than 2 weeks. They dither-locked the OMC, measured the FSR and finesse, and scanned for PZT resonances. The finesse was around 220-230, low compared to the design of ~300. They will assemble a careful procedure for disassembly, re-assembly, alignment and testing on the clean bench, and then installation, alignment and testing in the vacuum chamber.
  • Ben checked out the DC photodiodes. They are now functioning fine. The correct regulators are in place, and the diodes are healthy. He still want to bolster the protection of the photodiode elements, but he hasen't implemented anything yet.
  • Sam and Rana are getting the DC PDs mounted and testing their response and noise, using the DC PD satellite amplifier built by Ben.
  • Rob and Sam tested the tip-tilt PZT that was wired incorrectly; it shows no sign of being damaged. Rich, in consultation with Piezo-Jena engineers, established the correct wiring. Rob measured its dynamic range.
  • Sam is driving the tip-tilt PZT mirror with our old existing board. It seems to have a lot of excess noise at high frequencies. Under investigation. He and Darcy plan to implement dither locking of two angular degrees of freedom (along with the OMC length lock), soon. They will also test the 2nd tip-tilt PZT mirror to make sure it works correctly.
  • Ben reports that Alex is testing a computer with PCIE bus, at 32KHz, for us and LASTI. The order is in for components for interface boards, expected in on July 29.
  • Ben's QPD Whitening board PCB layout is finished except for a check to make sure all of the nets are complete. The board should go out in the next couple of days.
  • Rana and Sam are determining the noise and modulation requirements for the tip-tilt PZT and OMC PZT drivers.
  • Go reports that Evgueny is testing a new PPKTP crystal for the OPO. Go is preparing for this new OPO, including the mode matching and is ordering new lenses.
• PSL:
  • Steve reports that the MOPA head temperature has been stable for the last 2 weeks. He once again removed around 500 cc of water from the chiller... a great mystery.
  • Steve covered the top and two sides of the Mach Zehnder with a plexiglass box. The long-term drifting of the MZ servo has not changed much since then. Rana suggests building the other two sides.
• Electronics, controls, computers:
  • Bob plans on installing new RF distribution panels at the RF rack and the LSC rack, in the near future. It will take a couple of days and be somewhat disruptive.
  • Jenne is building a mount for an AM laser to be used for PD testing.
  • Osamu and David found a strange electronics noise on the PD2 photodetector; under investigation.
• Lab Infrastructure:
  • Steve and Bob replaced the forepump for TP2, and a new convectron gauge controller installed. At that point, all the Vacuum EPICS channels stopped appearing in EPICS and dataviewer. A DAQ reboot seems to have fixed the problem.
  • The presently-installed RGA filament has failed (the electron multiplier failed long ago). Steve gave a spare RGA head to Bob for baking. Bob will modify an oven head mount and bake it next week.

June 15, 2006

• Four new summer SURF students arrived On Monday: Jenne Driggers of U Washington, Darcy Barron of U Illinois, David Malling of Syracuse U, and Royal Reinecke of Caltech. They got SURF safety training, laser safety training, laser entrance eye exams, and a full safety walkthough of the 40m lab from Steve.
• IFO commissioning:
  • Dan continues to develop the WFS diagonalization script.
  • Dan and Osamu are working with SURF student Dave Malling on developing the noise budget model (and eventually, hunting and reducing the noise sources).
• IFO modeling:
  • Monica is back from Europe and back to developing her e2e model. Time to start applying it to issues in lock acquisition and DC readout.
• DC detection and vacuum squeezing development:
  • When Ben and Rob were unplugging cables on the DCPD head, a photodiode element blew. Ben swapped out the element, and the DCPD works fine. There was a small problem with the voltage regulators, which will be fixed tomorrow. Once it is fixed, We'll do an extensive reliability check, and protect the diodes wherever possible.
  • Ben is in the final stages of the QPD Whitening board PCB layout. It will go to PCB Express sometime soon.
  • Rich and Rob noted that the tip-tit PZTs from PiezoJena had different wiring than the ones we installed long ago. It might be that they were damaged when Rob and Sam powered them up. Ben and Rob worked with the existing (not new under-design) PZT driver board so we could test the PZTs to see if they were damaged. After much effort and testing, driving the mirrors with appropriate sinusoidal voltages, Rob determined the correct pinouts and safe voltage ranges, and determined that the PZTs were not damaged. Whew!
  • Ben's DC PD satellite box being is debugged by Rob and Ben.
  • Rob looked at the resonances in the mounted TT PZT mirror; they all appear to be above 550 Hz.
  • Bob and Ben are working on specifying and building the in-vac "pigtail" cabling for the DC readout beamline.
  • Go returned from the World Cup to discover his OPO temp controller was turned off. We have no idea why, it might have been an interaction with the janitor's mop. He fears the LiNbO3 crystal was damaged and replaced it with a spare (he and Evgueny plans to move to a newly developed PP-KTP crystal).
• PSL:
  • After turning down the temperature in the control room (where the laser chiller is), the laser head temperature and chiller temperature have been rock-solid stable. However, Steve still thinks the chiller water level is rising mysteriously. Under careful watch.
  • Steve covered the Mach Zehnder on the PSL table to reduce the airflow and thus acoustic noise and drift. So far, he has covered only the top, will do the sides next.
• Electronics, controls, computers:
  • Christian brought over a new windows PC for use by the students and lab personnel.
  • Bob is working on heliax strain relief on LSC rack and ISC tables, and labeling all the RF cables.
• Lab Infrastructure:
  • Steve hooked up accelerometers at 4 places around the lab. Still need to change the channel names in the DAQ to make them meaningful.
  • The RGA filament current is zero. Steve will replace the head or the filament.

June 8, 2006

• We had a meeting of the 40m Technical Advisory Committee Thursday morning, 6/8/06. Transparencies (G060280-00-R) are here, in ppt and pdf.
• Osamu gave a talk on the status of the 40m, and Monica gave a talk on the status of her e2e simulations of the 40m, at Elba. To be posted soon.
• On Monday, we expect four new summer SURF students to arrive: Jenne Driggers of U Washington, Darcy Barron of U Illinois, David Malling of Syracuse U, and Royal Reinecke of Caltech.
• IFO commissioning:
  • Rob and Rana continue to optimize sensing signals and lock acquisition procedures. The lock acquisition for the full dual-recycled FPMI now proceeds smoothly, and the last steps are being refined.
  • Rob and Rana have done much work on developing scripts for automating various lock acquisition steps.
  • The MC WFS is now working again, after the beamline was relocated to the AP table. Dan is working with Rana on optimizing the diagonalization of the ASC control plant.
• IFO modeling:
• DC detection and vacuum squeezing development:
  • Rob and Sam have completely aligned the DC readout beamline (on a table in air), locked the OMC, coarsely focused the OMMT, and seen signals on the DC PDs. They put a fiber-fed beam in place of the DC PD and ran the whole thing backwards; this is how they plan to align the system to the interferometer when it is installed into the vacuum system.
  • Rob set up a dither-lock system for alignment of the beam into the OMC, and closed the loop on one degree of freedom (of 4; pitch and yaw from two TT PZT mirrors); to do more we need more lockins (or the digital control system).
  • The DC readout beamline is now just about ready to be dismantled for clean/bake. We plan to install near the end of July.
  • The DC readout controls are in progress (Ben, Jay, Rolf, Rich). Ben finished and delivered the DCPD Satellite box so Rob can test it with the PD head. He's getting close to finishing the QPD Whitening board PCB layout. It should go to PCB Express sometime soon. On order: control computer, ADC and DAC cards, Refl Memory card. Being built: AA and AI interface boards. Design nearly complete: OMC PZT driver, TT PZT driver. We hope to have everything in place by the end of July.
  • From Jay: The AI circuit used for slower sample rate systems was tested for use in a 32K sample per second system. The circuit appears to perform adequately. No excess noise, slew rate limits or non-linearities were observed. This circuit will be used for the first systems installed. A different, higher frequency, circuit will be designed for the possibility of pushing the sample frequency to 64K or 128K samples per second. Similar tests will be performed on the present AA filter.
• PSL:
  • Our MOPA continues to have cooling problems, and the head temperature is inversely correlated with the output power. there is some evidence that the cooling water line is getting blocked, requiring rooting. Steve is consulting with Peter King on what to do.
  • Rana made a PSLwatch script in Perl. It uses the new 16 bit IFO State vector and manages the running of the autolocker scripts for the FSS, PMC, MZ, and eventually the ISS.
• Electronics, controls, computers:
  • Our Framebuilder RAID array was accidentally switched off, causing the DAQ to go down and controls to be very unhappy. It took a while (and Alex Ivanov's help) to find the problem. Meanwhile, all the control computers were rebooted.
  • Rana and Bob built a patch panel for the RF heliac cable distribution on the LSC rack. Now much better strain relief and organization.
  • Rana compared the frequency stability of some of the equipment in the lab.
  • Ben brought over the new timing distribution modules from LHO, that we will install sometime in the future.
  • Steve and Bob ran more video cable to/from the new location of the SP CCD camera and the video switch.
• Lab Infrastructure:

May 25, 2006

• IFO commissioning:
  • Rob and Rana continue to improve lock acquisition. Following more and more tweaking, signal quality improvements and lock acquisition code and script work, we are now firmly in a region where it is easy to acquire DRMI+2arms lock, and do a bunch of signal handoffs. Work on reducing the CARM offset to zero and returning to full arm resonance continues.
  • Dan continues to work on getting the MC WFS system back up after the move to the AP table (greatly shortening the beam path), 45 degree reorientation of the WFS heads, guoy phase recalculation, and power optimization to ensure WFS head safety without the need for fast shutters. He awaits the appropriate beamsplitters, on order; meanwhile, he'll get it working with imperfect beamsplitters, modify the matrix to handle the 45degree rotation, diagonalize the servo and get it recommissioned.
• IFO modeling:
  • Monica is in Europe, preparing a talk on her e2e simulation efforts for the Elba meeting.
• DC detection and vacuum squeezing development:
  • Rob and Sam have more-or-less completed the DC readout in-vac beamline alignment, on the bench, using an NPRO. All components are in place, including the PZT steering mirrors (not yet powered). The OMC dither-locks at 20 kHz, "like butter". They are working on improving the mode matching. The output mode matching telecope has a variety of problems, making the mode matching difficult, but not impossible. They have a plan for aligning and mode matching after installation in vacuum. They are now checking out the DC photodiodes, and will soon be ready to dismantle the entire thing for vacuum cleaning and baking. The system will then be reassembled and pre-aligned on a clean flow-bench in preparation for installation, currently scheduled for mid-July.
  • Rob, Sam, Ben, Rolf, Rich and Jay had a summit on the DC readout controls. Most of the required controls should be ready in time for installation.
  • Ben reports that the DCPD Satellite box's front and back panels will be here any day now.
  • Ben is finalizing the QPD whitening board design.
  • Bob tested some resistors for vacuum compatability, and found that the gold-bodied power resistors failed the RGA scan. We'll have to go with the ceramic bodied ones.
  • Jay built and tested a prototype for the tip-tilt PZT mirror driver needed for both the LASTI Ponderomotive and 40m DC readout controls.
  • Bob baked the conflats for the DC readout beamline ports (OMC reflected and transmitted, to QPDs) and gave them to Steve.
  • Go has been working on the stability of the SHG cavity. It's still susceptible to seismic and acoustic noise, but its stability is better than before. Work continues.
• PSL:
  • Steve and Dan, in the process of rewiring the PSL enclosure interlock, discovered that the interlock cable was flaky, causing some of the spurious interlock trips we've been experiencing lately. With help from Ben and Rich, they rewired the interlock so that it shutters the power amplifier rather than tripping the NPRO; much more gentle for the laser! As a bonus, the EPICS control of the power amplifier shutter now works.
  • The PSL head temperature is still observed to be fluctuating, but less than before. There have been no sharp spikes in HTEMP in the last few days. Steve continues to watch this closely.
  • Steve covered the PSL trip switch with a guard to prevent it from accidentally being pushed.
  • Rana noticed that reflection from the squeezer back to the PSL NPRO makes it much noisier, despite the presence of two Faraday isolators in the MOPA box. He blocked the light path and the noise went away. Go installed a Faraday isolator in his squeezer path, but it had little effect. He traced it to reflection from the OPO, not the SHG.
  • Steve noticed that the PSL FSS can lose lock (overnight) when IFO is left in LA mode; could this be due to the fast CM servo? Under investigation.
• Electronics, controls, computers:
• Lab Infrastructure:
  • We had some ~ 60 visitors to the lab on Saturday, as part of Caltech Alumni / Seminar day.
  • Steve reports that after the rains, the particle count is now back down to normal in the lab (it follows the outside particle count very well, unfortunately).

May 18, 2006

• IFO commissioning:
  • Rob, Rana and Sam have made considerable progress in lock acquisition. They have cleaned up many of the LSC PD signals, adjusting phases and gains, matching the 133 MHz and 199 MHz signals, etc. The DRMI now locks in only a few seconds.
  • With the DRMI locked on just the single demod signals (REFL33_I, REFL33_Q, and SP166_I) they can easily explore the full space of double-demod signal phases to optimize signals for each DRMI degree of freedom. They can then hand off the controls to the optimized double-demod signals in preparation for adding the arms.
  • Starting with a locked DRMI, they offset-lock the arms using the DC transmitted signals, and tune up the RF signals. They can then switch to the RF signals for DARM and zero that offset. The full IFO is now locking well and robustly and there is some confidence that the signals are well tuned up and understood. Up next is more CM servo work and reducing the CARM offset to zero to bring the carrier to full resonance in the arms. So far, they have been able to re-establish the digital common mode servo, and the fast AOM path was also briefly, apparrently successfully, enabled.
  • Rob developed the autoalign scripts, adding one for the SRM mirror. There is now an alignAll script which automates the alignment of the entire IFO.
  • The input MC reflected beamline and LSC diode is now on the AP table (much shorter beam path) and works well.
  • Dan and Rana have set up the MC WFSs on the AP table with 45degree mounts, and Dan is positioning lenses for the correct Guoy phases (modeled by Monica) and pickoffs for the correct amounts of power on each of the WFS heads.
• IFO modeling:
  • Osamu continues to develop his e2e model of 40m and AdLIGO, his analysis of the AdLIGO length sensing noise, and development (with Kentaro) of alternative AdLIGO length sensing schemes.
• DC detection and vacuum squeezing development:
  • Rob and Sam have aligned the in-vac DC readout beamline that will go in the vacuum, using a crystal laser, and they are now switching to an NPRO (borrowed from the TNI) so that they can lock the OMC with it. They are adding the PZT steering mirrors onto their alignment chain. They may be ready to disassemble the system for cleaning and baking as early as next week.
  • The DCPD Satellite box is now successfully tested by Ben, and is awaiting front and back panels to arrive early next week.
  • Rob is specifying the correct alignment of the DC PDs in their mounts so that Ben can do the final machining on the DCPD head mount.
  • Ben continues to work on the QPD whitening board layout.
  • Jay continues to work on the OMC PZT driver board and the steering mirror PZT driver board.
  • Bob got in some cable so we can pre-make the cabling for the DC readout controls.
  • Vacuum windows for the OMC reflected and transmitted beamlines are being baked out by Bob.
  • Go and Osamu continue to work on the squeezer. With 1.5 W into the SHG, it gets hot, and the SHG cavity becomes unstable. They lowered the oven temperature, and the cavity became stable, delivering up to 400 mW of green light continuously.
  • Go optimized the alignment of the OPO cavity using the seed beam, and did a temperature scan to find the largest amount of green light from the OPO operated as a SHG. He then aligned the pump beam into the OPO. He is now trying to get some parametric gain in the OPO. converting one green to two IR photons.
• PSL:
  • Several laser tripoffs in the last month have been traced to user error with the PSL enclosure. Everyone is now fully aware of the care that must be used to avoid having the PSL enclosure trip off the laser inadvertently.
  • Steve will arrange to have the PSL enclosure interlock changed so that it only shutters the MOPA laser, not trip off its power. This should help to avoid tripoffs that could damage the laser.
  • In addition to tripoffs, the MOPA head temperature wanders and the temperature servo is unable to stabilize it. Steve believes that this is due to temperature excursions that are accompanying the squeezer work being done on the PSL table, stressing the ability of the MOPA temperature servo to maintain head temperature stability. Steve will consult with Peter King and other experts on what to do about this.
• Electronics, controls, computers:
  • Rana replaced the ETM and ITM suspension controller violin mode 1st (2nd?) harmonic (884 Hz) notch filters with 4th order Elliptic Bandstop filters, which work much better.
  • Rana noted that many of our EPICS channels have names that are non-standard, look nothing like the naming conventions used at the sites, and don't match front-end names. He has renamed a large number of channels in the LSC system.
• Lab Infrastructure:
  • Steve notes that the particle counters are recording extremely high values, presumably due to the weather. He'll keep an eye on it and keep the HEPA filters running as often as possible.

May 11, 2006

• IFO commissioning:
  • Rana and Monica moved the MC reflected beamline to the AP table, shortening the beam patch dramatically. They reduced the light hitting the PDs to a safe level, eliminating the need for a fast shutter. They then re-phased the RF demodulation and are able to lock the MC. Steve moved a ccd camera over to look at the MCR light. The WFSs were also moved, but they require new cables which are being made by Bob, and new Guoy telescopes which are being designed by Monica. Dan is working on 45-degree brackets for holding the WFS boxes so that the PD is a +, not x, as was done at LLO, to reduce some of the coupling between pitch and yaw. The MC WFS will need to be re-diagonalized.
  • The small AP table is now a bit more crowded, (containing the AP RF, SP, and MCR beamlines, and the MCT beam going to the PSL table), but still has room for the DC readout QPDs and a path for the squeezed vacuum beamline. There are no longer any beamlines going all the way to the SP table.
• IFO modeling:
  • Monica is cleaning up her e2e simulation of the 40m. She succeeded in locking the simulation, using Osamu's seismic model. Optimization of sevo gains is in progress. She is also working on locking with DC signals at TRX and TRY.
  • Osamu is working on new length sensing schemes from ALIGO, using lower RF frequencies. It is harder to test such schemes at the 40m due to the short power recycling cavity, but he's trying to find such a scheme.
• DC detection and vacuum squeezing development:
  • Rob and Sam continue to put together and pre-align the DC readout beamline, in preparation for installation into the vacuum. We need to get a good NPRO for alignment and OMC locking.
  • Rob and Sam are working out an in-situ alignment procedure for the DC readout beamline after installation, using an NPRO (to be borrowed from somewhere) and a fiber feed.
  • Osamu touched up the MZ alignment. There's a lot more reflected light when in lock, because the modulation dpeths have been turned up to make IFO locking easier.
  • We tentatively plan to install the DC readout beamline in the vacuum in the 3rd week in July, when Rob comes back from his honeymoon. Bob can have everything cleaned and baked before then.
  • The DCPD/OMMT Satellite box has been stuffed by P.R. and Ben is testing it now. Everything seems to be working well, but Ben wants to make a comprehensive test procedure for it to make sure that it all works well.
  • Ben is working on the front panels for the DCPD/OMMT box. They should be done soon.
  • Ben is finishing the schematics and layout for the QPD Whitening board.
  • It is believed/hoped that Jay is working on the design of the steering mirror PZT driver board and OMC PZT driver board, as well as the new AA and DAC interface boards for the PCIX-based controls.
  • Ben made retaining clips for the OMMT mirror to replace the clips that came with it. The original clips were much too big, and covered a large portion of the mirror surface. The replacement clips are made from titanium, and are much smaller than the originals.
  • Go and Evgueny are optimizing the SHG; they are getting 620 MW of green light. "Seeing the green come out of nowhere is very beautiful." - Rana. This higher power makes it harder to acquire lock, and they are trying different methods to make it easier. They also installed the pump beam path into the OPO. They are generating green light in the OPA as well, to help guide the pump beam. They are also observing daily alignment drifts; this may be due to the MOPA trips (see below), or something else drifting upstream. They are investigating.
• PSL:
  • We started seeing laser head temperature hiccups around 2 weeks ago. Osamu thought they were consistent with PSL enclosure trip-offs. Steve did a controlled test, opening the PSL enclosure when it was interlocked, and saw precisely the same kind of hiccup. Either the enclosure is getting bumped, or an enclosure sensor/relay is getting flaky. Steve will try to locate a flaky sensor/relay. Meanwhile, the bumps should set off an alarm; this will be re-enabled. And an interlock trip-off should not trip off the MOPA. Steve will look for a softer way to provide the personnel protection.
• Electronics, controls, computers:
  • Dan ran a new heliac cable to the 33 MHz pockels cell, to replace a broken one.
  • Ben ordered 6 of the YAG-444-4AH (the RoHS compliant version) PDs from Pacer. Four are for the 40m, and two are destined for Livingston. Richard wrote that he had plenty, and didn't need any ordered for Hanford.
  • Front end test points continue to become unavailable for no apparent reason, requiring restarts of the c0daqawg processor.
  • We continue to have heaps of computer problems. Rob noticed that the RGAlogger hasn't been running for months, and the conlogger also needed to be restarted.
• Lab Infrastructure:

May 4, 2006

• IFO commissioning:
  • With much increased RF modulation and LO oscillator levels at the LSC rack (see below), DRMI locking is much easier and more robust. Re-optimization of loop gains in progress.
  • Rob and Rana have a new easier way to lock the DRMI (no arms) using only single-demod signals at the SP (no double-demod). This allows them to tune up the demod phases, offsets, and diagonalizing the plant for the double-demod signals is much easier, and the control can be handed over to those tuned-up signals before adding the arms.
  • Osamu continues to work on the noise budget with Monica and Dan. They have made tremendous progress. The code Osamu developed at LLO for improved measurement of the frequency noise has been brought over and implemented at the 40m. Most of the main noise sources for the 40m FPMI have been measured, including frequency noise, seismic, oplevs, MICH, calibrated DARM noise, Dark noise, frequency noise, seismic, oplevs, MICH, and their sum (total predicted DARM noise). Still need intensity noise; to measure, we need an analog output from the front end (AWG) to actuate on the PSL intensity. We will maybe use a spare, unused ASC (or IO suspension) Pentek DAC channel, Also missing is oscillator phase noise.
  • The noise budget indicates that the FPMI sensitivity is limited by dark noise above ~200 Hz, which will be improved when we increase the optical gain by locking the PRFPMI or DRFPMI. At that point, the noise budget code can measure the PRC and SRC loop noise. But we need to implement into the noise budget code a reliable model of the DRFPMI frequency dependence, including optical spring and RSE peaks.
  • The signal from the MC transmitted RFAM PD went away, possibly because an upstream optic was disturbed. The MC autolock script, which uses that signal to sense when the MC is in lock, was changed (by Rob) to use another PD.
• IFO modeling:
  • Monica is updating the seismic model in her e2e model of the 40m, using Osamu's model. It's 100x higher at 10 Hz. Now she's trying to lock with the much larger seismic noise.
  • Monica is developing a Bench model for AdVIRGO.
• DC detection and vacuum squeezing development:
  • Ben reports that the DCPD/OMMT Satellite box is being stuffed by PR and should be finished soon. Ben has started the front panels for that box.
  • Ben is finishing the QPD Whitening board layout for the DC readout beamline.
  • Ben gave Bob ten more resistors (for the in-vac DC PD) to bake and scan. They are an easily mountable wirewound type.
  • Ben machined the DCPD beamsplitter mount to allow for retaining clips to be screwed on, and Dan made and installed the retaining clips. It's not a neat solution, but it works.
  • Ben reports that Jay is working on the layout of the TT steering mirror PZT driver board and the OMC PZT driver board. Also, the PCIX-based computer, and ADC and DAC cards, are in hand. The ADC and DAC interface boards are designed and will be ordered soon.
  • Rob and Sam tried locking the OMC with the crystal laser borrowed from LLO, with no luck; it was too noisy. We need to find a spare NPRO.
  • Rob and Sam are unhappy with the mounting of the small curved mirror on the output MMT, and are thinking of alternative mounts using a new mirror.
  • Go and Evgueny installed the new OPO cavity, aligned and mode matched the seed beam, found the cavity modes, and optimized TEM00 transmission of seed beam. Next step is to align the pump (frequency-doubled) beam. They also mode matched into and out of the mode cleaner fiber they use for the local oscillator beam, and got 90% throughput. with Osamu, they are designing a single remote-insertable In-vac pickoff mirror for their squeezed vacuum.
• PSL:
  • Steve continues to monitor the PSL power. The internal MOPA AMPMON photodiode from Lightwave, strongly humidity dependent, was replaced by a new PD (thorlabs PDA55) external to the MOPA and calibrated to watts. We now have a reliable way to monitor the MOPA power before it goes through the PMC.
  • Steve and Rob noted that the PSL head temperature had a couple of "hiccups" in the last week; one was possibly due to optical feedback when a shutter was closed, but one or two others have no known cause, and everything fell out of lock. Being closely watched.
  • The MZ frequency noise is large, and the servo drifts into saturation easily. Steve will construct a plastic box around the MZ to protect it from acoustic noise and accidental bumps.
• Electronics, controls, computers:
  • Dan and Rana measured RF LO levels at LSC rack, and found lots of problems and inadequate LO levels.
  • Rob redid much of the RF distribution wiring, replacing the old frequency distribution boxes with Mini-Circuits ZFSC-3-13's. He then turned up the modulation at the RF distribution rack, and now there is much more LO and RF signal going into the LSC rack (although the 199 MHz is still a bit low). The Heliac for the 166MHz to the pockels cell has bad connector - to be fixed.
  • To aid in noise budget work, we need to have the seismic spectrum from the accelerometers at the vertex and at the ends, always available to the front end controls. Steve will set up the accelerometers and run the cables.
  • Our controls computers continue to be a mess, with many things not working on them. We will try to enlist help from solaris experts.
  • We need to prepare our PC computers for the onslaught of SURF students this summer.
• Lab Infrastructure:
  • Steve had the electricians install a 20m breaker from isolation transformer panel to the DCreadout rack.
  • Steve reports that a vacuum foreline gauge failed, to be replaced.

April 27, 2006

• IFO commissioning:
  • Osamu, Dan and Monica continue to develop the noise budget. Above 100 Hz, the ~ few * 1e-13 m/rtHz is entirely due to dark noise (electronics, etc). Oplev noise is significant in the 10-100 Hz range. Seismic noise dominated below 10 Hz, but there's a factor ~8 discrepancy in the absolute prediction; under investigation. Steve is moving the accelerometers around the lab to see how much variation there is in the spectrum. But the calibration of the accelerometers is in question.
  • Rob and Rana worked on lock acquisition. They have not yet managed to reduce the CARM offset to get to full power in the AdLIGO configuration, ever since the laser swap in November/December. They are working through all the changes and all the places where noise could be getting in.
• IFO modeling:
  • Monica and Osamu adjusted the demod phases in her e2e simulation of the 40m, and now the simulation goes to full power at the correct AdLIGO configuration.
• DC detection and vacuum squeezing development:
  • Rob and Sam have set up the output mode matching telescope (OMMT) and aligned it using a pair of autocollimators. The alignment is difficult, there aren't enough knobs to adjust; but they got it to work. We have a crystal laser from LLO that we will use to align the OMMT with respect to the ouptut mode cleaner and lock the OMC.
  • Ben finished stuffing and testing the DCPD in-vac electronics board and delivered it to the lab. It is now being set up for test.
  • Ben is about done drawing the schematic for the DC Detection QPD Whitening Chassis. He will start laying out the PCB sometime soon.
  • Go tweaked the input beam to his his second harmonic generator and was able to increase the green light from 300 mW to 350-400 mW.
  • Go has updated his conceptual layout for the squeezer and squeezing detector system on the PSL table.
  • Go is now putting together the beamlines that go into and come out of the OPO (which Evgueny is bringing next week). He designed three-lens mode-matching telescopes for the input and output beams. He's also discussing with Osamu and Steve about the placement of the pickoff mirrors needed to inject the squeezed vacuum into the asymmetric port.
• Electronics, controls, computers:
  • Osamu and Dan measured the noise at the RF output of the AP166 RFPD (our DARM signal), and found it to be ~ 50nV/rtHz, not unreasonably large.
  • Rob continues to make improvements and add useful features to the length sensing and lock acquisition front end code.
  • Rana and Dan measured the noise spectra on all 4 quadrants of both WFSs in the mode cleaner alignment servo; there are no obvious anomalies, and the alignment system seems to be working reasonably well. The fear that the WFSs were fried during some beamline work appears to be unfounded.
  • Rana and Sam rewired the RF signal generators to frequency lock them to each other properly. Now there should be no more ringing in the cable or standing waves. They also tied the 29.5 MHz IMCR oscillator to the others; So we should have less wander in the intermodulation AM produced by the Mach-Zender.
  • Rana and Ben figured out the installation wiring and module needs for the new MC Servo. We have most everything that we need to get started. Once we start, it should be one or two days of down-time until the new servo is operational.
  • Ben made a BNC switchbox with one BNC in, and four switchable BNC outputs, for Go.
  • The PSL still appears to be losing power, as reported by the PMC transmitted PD. Steve has installed a new PD before the PMC to monitor the power just as it exits the MOPA. It needs to be interfaced into DAQ/EPICS.
  • The PSL FSS loop drifted into saturation, which has happened several times before. Osamu traced this to a slowly falling amount of tight transmitted through the reference cavity. He tweaked the alignment and brought the transmitted power up a bit, but it is still only 2/3 of what it was 3 months ago. There is a sharp drop that occured on March 28. Under investigation.
• Lab Infrastructure:

April 20, 2006

• IFO commissioning:
  • Dan and Monica continue to develop the noise budget. Osamu, having learned about the noise budget program while at LLO for comissioning, is now working with them. They made a calibrated noise spectrum with the FPMI at the DARM port; noise level is at ~ few * 1e-13 m/rtHz. Noise hunting is just beginning...
  • Rob and Osamu continue to work on alignment and locking scripts, working with the new length control front end code changes.
• IFO modeling:
  • Monica can now keep lock on all 5 degrees of freedom in her e2e simulation of the 40m, even with radiation pressure on (she turned up the gain on the MICH loop). She's working on improving the stability. She's developing the AdVIRGO model.
• DC detection and vacuum squeezing development:
  • Rob is working on the setup and alignment of the full DC readout in-vac beamline.
  • Ben got the DCPD in-vac electronics board back from PCB Express. He will stuff and test one soon.
  • Ben also got back the DCPD Satellite box PCB, and will also stuff and test this board soon. So far, it fits in the box. That's a good start.
  • Ben is drawing the schematic for the DC Detection QPD Whitening Chassis. He should be able to start laying out the PCB sometime next week.
  • Go and Osamu are designing the pickoffs for injecting sqeezed vacuum.
  • Go reports that his second harmonic generator is putting out 300 mW of green light, with ~ 2 w input pump beam. The intensity is stable to ~ 5%. Evgueny is coming next week with the new OPA.
  • To aid in the in-vac installation of the DC readout and squeezed vacuum components, Steve will assemble a full list of in-vac cabling.
• Electronics, controls, computers:
  • Rob, Rana and Monica found that the frequency setting on one of our RF signal generators (the 199 MHz) was set wrong by 20 Hz, making the 199 MHz demod signals completely wrong. This must have happened since the last time any serious locking work was done, ~ 1 month ago.
  • Rob and Rana found some problems with the common mode servo board which were producing wierdness in the transfer functions. They changed some resistors and the board now behaves as expected.
  • Ben and Rana brought the MC Servo, CM Servo and DCPD parts over to the 40m on Wednesday. They plan to get together on Friday to figure out the installation wiring and module needs. The plan is to install the MC servo board by early next week, and then possibly move on to install the new CM servo board.
  • Ben is preparing an order for some YAG-444-4A WFS QPDs to replace two broken MC WFS PDs, and to also have some spares for the 40m, and the sites.
  • Steve notes that the PSL appears to be losing power, 10% over the last 24 days. not clear if it's in PSL or downstream. He will put another independent PD after the PSL but before anything else.
  • Steve continues to make plans for a mezzanine optical table in the PSL enclosure as a new home for the MCR beamline (much reduced path length compared to the present).
  • Steve got a quote for the repair of our crystal laser: $1000. Meanwhile, we will use the autocollimator and hold off on this repair until we're sure we need it. Steve and Mike will set up the autocollimator for full alignment of the DC readout beamline.
  • Larry fixed the root password on our new operator's computer, op540m, and also fixed su (more precisely, sudo) on op440m. We're working down the very long list of things that don't work on our operator computers.
• Lab Infrastructure:

April 13, 2006

• IFO commissioning:
  • Monica and Dan continue to work on the noise budget, including contributions from seismic noise, optical levers, mode cleaner angular noise, frequency noise, and electronics noise (coil drivers, DACs, etc). They tried taking a calibrated noise spectrum with the FPMI, but the AP166 RFPD signal was too noisy. Under investigation.
• IFO modeling:
  • Monica continues to work on improving her e2e simulation of the 40m/AdLIGO configuration, with full 5 length DOF control system. She finds that the system won't stay in lock with all 5 DOFs under control, with radiation pressure on all mirrors. Under investigation.
• DC detection and vacuum squeezing development:
  • Ben is managing the acquisition of control electronics to be bought/made/designed for the DC readout system, including custom boards (TT-PZT steering mirror driver board, DCPD in-vac board, DCPD satellite board, OMC driver board), cables, and PCIX-based ADCs/DACs. It looks like we could get everything together by early June.
  • Jay is working on the design of the TT-PZT steering mirror driver board.
  • Ben sent out DCPD in-vac board and DCPD satellite board out to PCBexpress for fab. They should be back next week for stuffing.
  • Jay procured and tested the QPDs for the DC readout alignment control system.
  • Steve and Go put together a new open frame equipment rack for the squeezer electronics, taking power from a new circuit (unconditioned power).
  • Go got his second harmonic generator (SHG) working. He replaced the shaky 6-axis mount with a 3-axis positioner, realigned and mode matched, and then easily locked te SHG. He swept the temperature from 81 to 92degC for phase matching. The cavity lock is stable for hours. At maximum, he's getting 400-500 mW of green light out of 2 W input beam. He expects up to 600 mW with improved mode matching and alignment. The intensity stability is ~1-2% at peak output.
• Electronics, controls, computers:
  • Rob made major changes to the length control and lock acquisition front end code, including: (a) fast input matrix Ramping; (b) "bait & switch" which allows one to switch from a demodulated diode signal to some calculated signal, during lock acquisition; (c) a new lock acquisition mode word which summarizes the state of the most relevant lock acquisition settings with a set of bits that can be turned on or off via scripts and enabled wuth a user request button; (d) the "moving zero" filter code that compensates for the optical spring peak as the CARM offset is reduced is now more flexible; (e) and the code is generally cleaner and easier to read.
  • The MC WFS QPDs are probably damaged from too much light power. Ben went looking for spare QPDs, but found none. He plans to order a bunch for the 40m and the rest of the lab; 8 week lead time.
  • A breaker was accidentally switched on the electronics rack housing the digital suspension system. After some struggles, the system was restored.
• Lab Infrastructure:
  • Steve is thinking of how to add a mezzanine optical table in the PSL enclosure to provide more table space for the mode cleaner reflected, mode cleaner transmitted, AP and SP beamlines.
  • Osamu and Rana are at LLO to help with commissioning activities.
  • Bob is at MIT to help with the quad suspension controls prototype at LASTI.

April 6, 2006

• IFO commissioning:
  • Osamu and Rana are at LLO to help with commissioning activities.
  • Dan is diagnosing the MC WFS's, which Osamu reported to be noisier than before and which Dan suspects were fried by the beam during some work a few weeks ago. He has measured OLTFs of the MC alignment servos, which look ok but he doesn't know what they should be compared to. He and Ben will replace the WFS photodiodes with new ones to see if there's a difference. Ben is currently looking for the stash of YAG444A Quad photodiodes that they used to have at Wilson House.
  • Dan and Monica continue to develop the noise budget software. They are in the process of calibrating the DARM signal using a simple Michelson.
• IFO modeling:
  • Monica's e2e model of the full 40m control plant seems to work well. She has measured the OLTFs for all five length control loops. The CARM loop is derived from the signals at POX and POY, and there is an offset (~ 1 pm) from full arm resonance. She is working on undertanding and reducing this. She can now lock all five degrees of freedom, and it maintains lock throughout the 2-second simulation. It takes a good fraction of a second for the fields and the mirrors to reach equilibrium; under investigation.
• DC detection and vacuum squeezing development:
  • Rob and Sam glued the OMC PZT-mounted mirror bracket on to the OMC, and locked the OMC to a pickoff beam from the main IFO REFL port. They measured the finesse to be ~ 500 (design is ~ 100); the discrepancy may be due to the polarization of the input beam being wrong. They were not able to lock the OMC to the beam from the small crystal laser, due to too much noise on the laser.
  • That laser was sent out for evaluation/repair by Steve.
  • Next, Rob will assemble the output mode matching telescope (OMMT) and align the whole DC readout beamline (PZT steering mirrors, OMMT, OMC, DCPD) together.
  • Jay has completed a draft of the DC readout control system layout and comments have been received. Design of the system and the boards needed for controls should resume next week.
  • Ben reports that the layout of the DCPD Satellite board is nearly complete. He's just adding some annotation and test points, and it will be sent out to PCB Express for fab.
  • Go assembled his second harmonic generator (SHG) in its oven and enclosure, heated it up, and mode matched the PSL-picked-off light onto it. He observes green light. He is having difficulty locking the SHG cavity because of an oscillation somewhere. He put in a notch at 5 kHz, but a big 120 Hz oscillation remains; and there's lots of acoustic noise from the HEPA filter. Work in progress to lock the SHG cavity.
• Electronics, controls, computers:
  • Ben replaced the 166MHz I&Q demod board in the LSC rack. It was used for the AP 166 signal, but was very noisy. Ben was able to find anything wrong. Meanwhile, the SP 166 demod board was put in its place, so Ben returned this board to the SP 166 slot in the rack. Later, he will do a quick measurement of conversion loss, and check it against the other 166MHz board, and the Minicircuits spec.
  • Matt wrote a suite of IFO Test perl scripts for doing automated IFO diagnostics. He tested them at the 40m lab, then he and Vuk went off to the sites to install them. Rana exercised a mode cleaner test script at the 40m, and Rob is writing a new RestoreXarm script that makes use of the new suite. Rana and Rob plan to develop more scripts to test all the various LSC/ASC systems (oplevs, WFSs, IFO length control loops in various configurations).
  • Rob rewrote much of the LSC lock acquisition code to make use of a state vector in which the various elements of the IFO configuration are encoded in bits. This replaces the clunky " Mode 1, Mode 2" of lock acquisition. Rob augmented Matt's IFO Test suite with some code that reads and changes the bits in a lock acquisition state vector, to make it easier to test, diagnose, and drive lock acquisition sequences from scripts. Also, the input matrix code is now part of the LSC code instead of being a separate routine. All of these changes are still under test.
  • Rana installed a new version of StripTool, and used it to monitor the slow drift of the PSL power. This is basically a replacement for having a fancy mechanical stripchart recorder.
  • Steve noticed that the PSL FSS has been drifting out of dynamic range, requiring some jiggling of the sliders. He is documenting the symptoms to diagnose the problem.
  • Our operator control computers have various kinds of problems, and the whole system needs to be fixed up. Rob is compiling a list of known problems in the 40m wiki.
• Lab Infrastructure:
  • The maglev turbopump failed on Tuesday, and the pressure in the main vacuum rose to 2.3 mtorr before Steve noticed it. This underscores the need for a working alarm system. Steve doesn't know why it failed, but he was able to restart it. The software interlocks closed the main valve and protected the RGA. Meanwhile, the new replacement maglev is ready to be swapped in when needed.
  • Steve has called in electricians to add more AC outlets around the lab (especially near the PSL) for auxilliary equipment (such as Go's squeezer electronics) without having to run long extension cords and power strips. This is the last action item that came up during the safety inspection in January.
  • Bob will be at MIT next week to help with the quad suspension controls prototype at LASTI.

March 30, 2006

• IFO commissioning:
  • Go and Osamu are measuring the noise spectrum of FPMI at high frequency (~1 MHz). The demod board was very noisy, so they switched to another mixer. The noise is at a level of ~1e-11 m/rtHz, compared with shot noise expectation of 1e-15 m/rtHz. This appears to be due to the photodetector electronics. Osamu will follow up. Meanwhile, they will try to measure the noise in a DC readout scheme suggested by Rob.
  • Monica and Dan continue to develop the noise budget software for the 40m environment. They have measured the oplev couplings and transfer functions with the FPMI locked, and can plot the oplev noise budget traces. They are now working on the MC WFS contributions. They are modifying the XML templates for many, many channels. They will soon take a calibrated noise spectrum in the FPMI configuration.
  • Last week the MC was very unstable. Osamu found the MCR beam was clipped at the beam tube newly installed by Steve. The WFS signals had big offsets that could not be zeroed electronically. There might still be some clipping of the beam, or the WFS itself might have been damaged. Dan will follow up on the potentially damaged WFS photodiodes.
• IFO modeling:
  • Monica continues to develop her e2e simulation of the POX and POY loops, reducing the offset that appears in the full DRFPMI configuration. She is now starting an AdIRGO model in e2e.
  • Rob announces the first release of Optickle, a new matlab-based frequency-domain IFO simulation tool that includes quantum noise, radiation pressure, DC and RF readout, simple user interface, and lots of example configuration files. It can calculate TFs from any point to any other point in the simulation. He already has used it to note that the near-zero terms in the length control matrix from DARM to any double-demod short-DOF port get significantly larger when the radiation-pressure-induced optical spring is included. He is now handing the code off to Osamu to study the quantum shot noise in all the control loops, for AdLIGO. See Rob to obtain the code.
• DC detection and vacuum squeezing development:
  • Sam and Rob mounted the mirrors on the OMC body, and mounted the curved mirror on the PZT with a few dabs of epoxy. They then tried to see FP resonances with a small crystal laser, but the laser power was drifting, and there was mode hopping. So instead they directed some of the main laser beam from the IFO reflected port beam (which was already on the table, ~10 mW) into the OMC. They see nice FP flashes. They tried locking the cavity with the PZT-actuated mirror, using offset locking. However, the curved mirror was shaking too much (the epoxy was not yet fully cured), and the amplifier was insufficient. They then tried dither locking, but had similar problems. So now they are waiting to fully cure the glue joints securing the PZT to the bracket and the bracket onto the OMC body; waiting 24-72 hours.
  • Rob will set up the output MMT next, aligning it with a HeNe laser. Sam will set up the steering PZT mirrors and test to see if the mounts are sufficiently rigid.
  • The driver electronics for the tip-tilt PZT steering mirrors are still under design. Sam is deriving noise requirements, and Jay is designing a board with 2*2 PZT drivers, incorporating read-back of the tip-tilt strain gauges and a fast "shutter" level.
  • Jay has finished a detailed control electronics schematic for the DC readout, based on digital controls with a PCIX-based system.
  • The DC readout control software will be developed in "Bork-space", a new suite of power tools by Rolf to design the controls in simulink and then have it automatically generate the front-end code, EPICS screens and databases, etc. This will be fun!
  • Ben is almost finished laying out the PCB for the DCPD Interface board. He has met with Jay to make sure that the board is the correct design to fit with Jay's system schematics.
  • Jay says that we can use oplev PDs and interface boards instead of QPDs for monitoring the DC readout beams; they will be less noisy.
  • Steve, Osamu and Go installed a pickoff beam path on the PSL table to direct ~ 2.7 W of laser light (which was being dumped) onto Go's squeezer setup. They also shifted the MCT beamline (which was on the NW corner of the PSL table) over to make more room, and removed the Tropel. go installed the SHG optics, assembled the SHG cavity in its oven with a curved output coupling mirror. He also drew up version 2 of the squeezer layout. Parts for the SHG enclosure (for safety and acoustic noise) are being machined at the Caltech machine shop.
• Electronics, controls, computers:
  • Dan and Ben did a comprehensive survey of the fast DAQ channels. from the PSL/IO DAQ interface board. Ben is assembling a list of which interface board channels go to what signal names on the dataviewer screen. This list has slightly changed from the last time that he did this survey. He also compiled a list of the signal conditioning on the two Generic DAQ Interface boards. He will write up a table showing each channel's electronics from beginning to end.
  • Dan worked with Ben to debug the AP 166 demod board, studying how the LO levels affect the noise.
• Lab Infrastructure:
  • Steve measured the tranmission of IR light through our new laser safety glasses: Laservision's white frame "LC542" only for 1064nm OD 5+, and Bacou-Dalloz's blues-gray frame "31-20200" for 532 & 1064nm OD 5+. These have dielectric coating and they reflect the laser beam. He checked this at different incident values with 2.5W ~2mm beam diameter.

March 23, 2006

• Keisuke Goda ("Go") from Nergis' Quantum Optics group at MIT will be visiting for the next ~ 6 months. Please make him welcome!
• Fumiko Kawazoe visited from NAOJ in Japan and gave a talk on the status of her 4m RSE prototype interferometer. She has locked the FPMI and is moving on the PR and RSE.
• Osamu presented a status report on 40m progress at the LSC technical plenary session.
• IFO commissioning:
  • Because of some jostling, the beam exiting the MCR beamline was misaligned. Dan realigned things and re-established MC lock.
• IFO modeling:
  • Monica now has all of the length control loops implemented and tested, except for PRC (which comes next) in her e2e simulation of the full 40m/AdLIGO optical configuration. She fixed the problem with her POX and POY loops and the OLTF now looks fine.
  • Osamu presented some calculations by him and Kentaro on AdLIGO shot noise sensitivity and loop couplings including exploration of various different sensing schemes, at the LSC session on AdLIGO ISC.
• DC detection and vacuum squeezing development:
  • Rob, Sam, Ben and Jay have been working out the details of the DC readout controls. Jay is drawing up detailed schematics for the digital controls, based on PCIX technology (instead of VME).
  • Rob and Sam glued the OMC PZT to its bracket and the small mirror to the PZT.
  • Go installed some of the components of the vacuum squeezer onto the PSL table, and installed pickoff mirrors to direct the dumped light from the meain PSL beam over to his setup.
  • Go continues to develop his conceptual layout of the squeezer system on the PSL table.
• Electronics, controls, computers:
  • Rana found that the PSL/IO Anti-Aliasing chassis from the PSL rack was almost useless; many of the signals were railed or extremely noisy. Dan is desoldering all the filter chips and replacing them each with a single jumper wire.
  • Rana changed the DAQ datarate for the X and Y arm QPDs from 2048 to 16384. He then discovered an aliasing problem that is supposed to be performed by low-pass filtering in the front end code. He asked Rolf to fix it.
  • Rob and Rana re-commissioned the AO path of the Common Mode servo, using REFL_DC (from SP33) as the signal.
  • Rob made some changes to the LSC code, including the addition of the REFL_DC signal, and a re-implementation of fast input matrix ramping without incuring cpu slow-down.
  • Ben continues testing of the noisy I&Q Demod board for ASPD 166MHz.
• Lab Infrastructure:
  • Steve finished the installation of protective beam tubes on all the high-powered beamlines in the PSL enclosure, including the MCR beamline.
  • Steve put the PSL enclosure HEPA filters on a Variac and is running them at 40% of 120V. It is much quieter and particle counts are zero for 0.3 and 0.5 micron at the north west corner of psl optical table.
  • Steve and Go have assembled several pairs of laser goggles which work in both the IR and the green, for protection when working with the new second-harmonic-generator (SHG).

March 16, 2006

• IFO commissioning:
  • Rana, Rob and Sam made some progress in restoring full lock. They can reliably lock the DRMI+2ARMs interferometer in the anti-spring state with offset-CARM; acquisition was taking 1-3 minutes, and lasting until they broke something. They could engage the digital path of the CM servo, but not the AO path. They continue to be stymied by LSC signal electronics problems, (see electronics section below) and are working through them.
  • Dan and Monica are measuring the couplings from various channels into DARM, including secondary loops, mirror alignments, oplev loops, etc. They have determined suitable excitation levels. Now they just need to lock the IFO and start measuring couplings. They have been thwarted so far by various problems in the scripts, AWG, test points, etc, and are working through the problems.
  • Osamu and Steve adjusted the alignment and minimized the reflected light in the Mach Zehnder and input mode cleaner systems.
  • Alan is circulating a new draft of the 40m optical response paper; hope to submit for LSC review in the next couple of weeks.
• IFO modeling:
  • Monica continues to "commission" the length control servos in her e2e simulation of the full 40m/AdLIGO optical configuration. She is measuring the OLTFs of the different loops with DARM on and off and radiation pressure on and off. MICH looks sensible and she is moving on to SRC. She found the problem with her POX and POY loops that made single arm locking difficult, and is following up on it.
  • Rob is back from NAOJ Japan, where he worked on a prototype displacement noise free interferometer, with Seiji, Yanbei and Keiko. Two publications are in the works.
  • Rob has implemented RF readout in his Optickle modeling tool (matlab-based, with full implementation of the Corbitt et al quantum formalism).
  • Osamu is working on AdLIGO noise calculations.
• DC detection development:
  • Ben talked with Todd and procured some of the electronics that are needed for the DC readout scheme. The OpLev chain is preferred by Rana for DC detection alignment, but none of the boards are here. He spoke with Jay, who felt that we should all sit down and design the system, then decide which electronics we should make based on our requirements. Ben will get a meeting together with Rana, Sam, and Jay to make this happen soon.
  • Sam added an optical chopper onto his DC readout alignment control test beamline, to diagnose the anomalous loss of phase. He traced most of it to the oplev QPD readout board which has two 800 Hz butterworth LPFs. He's thinking about how to fix this in a modified board for the final system, in consultation with Jay and Rana.
• Electronics, controls, computers:
  • Ben, Dan and Rana have been looking at the noisy LSC demod boards. Ben and Rana measured the noise with different LO input levels, and the amount of RF leakage, and discovered some anomalies. The demod board for ASPD 166MHz signal appears to be a factor of 10 more noisy than the other boards. Ben is testing it at the Wilson House to determine if the fault lies in the board. Meanwhile, Rana is finding various problems with both the demod boards and the PD DC Interface Boards, which are crucial for lock acquisition.
  • Ben entered the LSC Whitening Interface Board connections at 1X2-2-12 into the ilog.
• Lab Infrastructure:
  • Bob cleaned the beam tubes for the PSL enclosure, and Steve is installing them. Should be done by next week.
  • Steve is experimenting with turning off the HEPA filter in the PSL enclosure during commissioning to see what effect is has on the Mach Zehnder drift and the laser noise.

March 9, 2006

• IFO commissioning:
  • Dan and Monica continue to develop the noise budget measurements and code. They tried to measure the contribution from shot noise, but find that they are limited by electronics noise. Rana will fix the PD interface boards to reduce this. They are now measuring the couplings of other loops and oplevs to DARM.
• IFO modeling:
  • The magnitude and phase of the optical transfer functions taken back in October by Osamu while varying the CARM offset are in beautiful agreement with the B&C predictions, for three different offsets and arm powers. The problem with understanding the phase turned out to be a bug in Alan's interpolation of the data. Rob gets good agreement with Optickle. New draft of the 40m optical response function paper has been prepared.
  • Monica continues to debug the control loops in her e2e simulation. She's taking transfer functions, tuning demod phases, etc. She measured the MICH OLTF, and Hiro compared it with a simulation from Twiddle; they appear to agree well.
• DC detection development:
  • Ben continues work on the DCPD satellite box. He talked to Rob and got the specs for the filters and the design is mostly finalized. Some questions remain on the placement of one of the filters, either in the vacuum nipple, or in the satellite box. It should be resolved shortly.
  • Ben has procured some of the electronics that are needed for the DC readout scheme, including a PMC servo board containing the PZT driver we will use to lock the output mode cleaner. The QPDs that were hard to get may be replaced with an OpLev chain, instead. Ben will talk with Todd again.
  • The beamsplitter for the DCPD detector was coated incorrectly. Steve has received a correctly-coated replacement.
• Electronics, controls, computers:
  • Osamu noted that the DARM signal from the AP 166 demod board has a lot of electronic noise, presumably from the mixer. He and Rana will look at different mixers, or a better RF amplifier upstream.
  • Ben did a survey of DC signals from RFPDs connected to the LSC Interface Board at 1X2-2-12. He traced all of the connections, and found out where they go, and will enter the list into the ilog.
  • New passwords have been implemented for the controls and root accounts on all the 40m controls computers (which are behind a gateway that requires yet different passwords).
• Lab Infrastructure:
  • The carpentry work in the 40m entrance alcove is done. Bob set up a drill press there, and will post safety signs.
  • Steve replaced a Turbopump TP2 foreline pressure rose to 200mT after 10,000hrs of running. Steve replaced the forepump with a rebuilt unit. He made a mistake that caused the main vacuum pressure to jump to 1 mTorr, but he quickly recovered from it. He also rerouted a vacuum exhaust line.
  • Steve is assembling the safety beam tubes for the MC reflected beamline and the PSL high-power beamline, in the PSL enclosure.
  • Dan took shifts at LLO and went to Mardi Gras.

March 2, 2006

• IFO commissioning:
  • Osamu and Noriyasu worked on re-acquiring full lock, but ran into many problems, associated with the new laser (different optical gains, etc), and changes made to the length control front-end code. Osamu found a bug in the code and worked around it. He was able to lock the FPMI and also move to CARM/DARM, but when trying full RSE lock the system becomes unstable as the CARM offset is brought to zero. Under study.
  • Alan, Osamu and Rob are trying to understand the magnitude and phase of the optical transfer functions taken back in October by Osamu while varying the CARM offset (and therefore, the power buildup and optical spring). The magnitude of the TFs agree beautifully with the theory expectations and with Rob's Optickle simulation; but the phase is a little bit off (consistent with something like a small negative delay). Rob thought it might be a phase difference between DARM_IN1 and PD1Q, but Osamu measured that and saw no phase difference at all. Under investigation.
  • Monica continues to commission the various pieces of the noise budget estimation code. Ben wired the AS166 DC signal into EPICS for use in computing the shot noise contribution to the noise budget. Monica modified the code to make the shot noise budget semi-automatically. She's now working on measuring the coil driver / daq noise in the test mass suspensions.
• IFO modeling:
  • Monica continues to develop and debug the control loops in here e2e model of the 40m. To aid in debugging, she is calculating the OLTF of the loops; so far, she has done DARM and MICH, and is working to understand the results.
• DC detection development:
  • Ben is working on the DCPD satellite box. He has talked to Rob and gotten the specs for the filters, so now the design can be finalized.
  • The beamsplitter for the DCPD detector was coated incorrectly, so Steve sent it back to CVI for recoating.
  • Steve ordered more vented screws for the OMC.
  • Sam and Jay agreed on some modifications to the driver board for the PZT steering mirrors in the in-vac DC readout beamline. Sam continued to investigate the resonances in the PZT mirrors. In the DLC mounts, there are several resonances below 1 kHz, including one as low as 150 Hz, limiting the bandwidth of the proposed ASC servo. Sam mounted mirror on bricks, and found no resonances below 650 Hz. So, we need better mounts!
  • Jay and Rolf propose to use the new PCIX-based readout architecture for the DC readout data acquisition and controls. Jay will draw up a detailed plan (building on Rob's block diagram) in his spare time. We can hope to have a system in place in a couple of months.
• Electronics, controls, computers:
• Lab Infrastructure:
  • Steve is building a beam tube for the MCR beam, and has sent it out for anodizing. Steve is also building beam tubes for the high power beam path on the PSL table.
  • A troop of Pasadena boy scouts visited the lab on Friday.

February 23, 2006

• IFO commissioning:
  • Osamu and Noriyasu measured the calibrated noise spectrum of the XARM, and the Fabry-Perot Michelson (FPMI). Measuring the XARM with POX 33 MHz, the noise floor is at ~ few times 10^-15m/rtHz with tuned up whitening gain. The FPMI, measured with AP at 166 MHz, is much noisier: the floor is at ~ 10^-13 m/rtHz. This is apparently due to noise at the demodulator. He will work on reducing this, and/or switch to demodulation at the AP at 33 MHz.
  • Osamu and Noriyasu refined the FPMI autolock script and tested it. They also are locking the DRMI routinely. Full lock in RSE remains a bit elusive; much work is required to reoptimize all the servos with the new laser. In progress.
  • Osamu and Noriyasu were stymied by problems in using dtt to take spectra and transfer functions (timeouts, etc). Probably due to time synch problems. We need Alex to fix this.
  • Monica is working on measuring the shot noise contribution to the DARM noise. She needs more DC signals into the LSC system and will work with Ben on that. She's also working on measuring the optical lever noise. She will try to measure the noise at the AP, in the FPMI or PRFPMI configuration, and calibrate to m/rtHz using Osamu's method.
  • We have a new A/C unit in the control room, which is now cooling the laser chiller effectively (as well as making the room more comfortable for humans). The MOPA temperature servo can now do its thing, and MOPA_HTEMP shows constant temperature at 18.55+-0.03 degree. The chiller water stays at 20deg. So we're out of the woods with the laser temperature fluctuations, which were making it impossible to operate the interferometer.
• IFO modeling:
  • Monica's e2e model of the full 40m configuration now holds lock for many seconds, using Osamu's seismic noise block.
• DC detection development:
  • Steve has sent the new in-vac mounts for the DC readout beamline to the shop for remachining.
  • Ben reports that his in-vac photodiode mount for the DCPD is fully fabricated, including the PEEK parts. He machined several of the parts himself in order to make them in an expedient manner. The entire mechanical assembly is now finished.
  • Ben is working on the DCPD satellite box schematic and the testing of the head electronics.
  • Ben is working on a list of the remaining electronics boards needed for the DC readout controls.
  • Sam Waldman is working on the DC readout in-vac steering mirror PZT driver schematic. He has ideas for many improvements on the current design, and will work with Jay to make a board. He needs a laser to test is ASC system (and we also need one to test the OMC and other aspects of the in-vac DC readout beamline).
• Electronics, controls, computers:
  • Electricians installed two new cable trays to route AC cords, requiring less long extension cords. This was an issue raised during the safety inspection.
  • Alex configured the network and network yellow pages service on op540m, so it can be used now to run Epics screens, etc.
  • The EPICS vacuum controls still have problems that need Jay's attention.
• Lab Infrastructure:
  • Royal Reinecke will be SURFing at the 40m this summer. She is starting by characterizing Mike Smith's long awaited oplev telescope. She just went through the standard 40m safety training with Steve.
  • Steve notes that we don't have adequate monitoring of the temperature in all the relevant locations (control room, IFO hall, PSL enclosure, and even in the vacuum chamber). This requires some attention to PEM-type channels in EPICS.
  • Steve continues to research cooling solutions for the new maglev turbopump.

February 16, 2006

• IFO commissioning:
  • Our replacement MOPA pulls more current and runs much hotter than the old one. The chiller is in the control room, and the air conditioning in the control room is broken. So the chiller is working too hard, and can't keep the MOPA cold enough. The PID servo loop has no headroom, and the MOPA temperature (and power) is fluctuating daily. This causes the interferometer control loop optical gain to vary, making it hard to acquire and keep lock. Attempts by Steve and Peter King to turn down the MOPA current have not been successful. We're scrambling to get the air conditioning working in the control room, to give the chiller some help. Meanwhile, we're worried that we do not have a spare MOPA....
  • Dan and Monica are making lots of progress on getting the noise budget code working at the 40m. The program now runs, but does not yet produce meaningful predictions. They will consult with Rana when he gets back next week.
  • Osamu and Noriyasu are reviewing the digital controls infrastructure and code; Noriyasu plans to build a similar system for TAMA. They have measured the ADC noise and DAC noise; they look pretty reasonable. the Pentek ADC has a broad peak (the Pentek "hum") and large 60 Hz / harmonics peaks. The LSC whitening board DC gain is too low to see detector noise; we need to implement a low-noise "detection mode" electronics chain as is done at the sites.
  • Rob, Osamu and Noriyasu have been working on acquiring full lock, but found a bunch of new problems, preventing long locks. Part of the problem might be the instability of the PSL due to the chiller problems mentioned above. It looks like each of the servos need careful retuning to be able to once again acheive robust lock. Osamu will work on this while Rob is away visiting Seiji in Japan next week.
  • Once robust lock is again acheived, Osamu plans to measure a calibrated noise spectrum, including high frequencies (MHz). He will also explore the RF and HOM content of the AP beam, in preparation for DC readout studies.
  • Mike Smith has a prototype optical lever telescope. We will bring it to the 40m and test it soon.
• IFO modeling:
  • Monica continues work on her e2e control model for the 40m. The DARM loop won't hold lock for more than a fraction of a second; she and Osamu believe that this is because the other degrees of freedom also need to be controlled well, so she's now implementing all of them. She cannot take a meaningful transfer function without the servos working well for all five length degrees of freedom.
• DC detection development:
  • Rob has drawn up a fairly detailed layout for the DC readout controls, including all cables, electronics modules, front-end and EPICS channels. He's got two stages: in the first, the oputput mode cleaner servo will be an analog system using an SR830 lockin. In stage 2, the servo will be all digital, using a prototype new PCI-X computer. He's working with Ben to design and/or acquire the electronics that we don't already have in hand, including the DCPD satellite box and the OMC PZT driver / summing box.
  • Sam Waldman continues the development of the tip-tilt PZT driver boards.
  • Ben reports that the DCPD boards are functioning well. The last rev of that board will be sent out next week for fab.
  • Ben got the in-vac photodiode mount for the DCPD back from the machine shop, and it looks great. He has delivered the PEEK for some of the remaining parts, and they are being manufactured currently. There are four small parts that remain, and he will machine those next week.
  • Rob and Dan moved the QPD and camera for the SP beamline to another place on the SP ISC table, to make room for staging and pre-aligning the DC readout beamline.
• Electronics, controls, computers:
  • Bryan Barr continues to work on a publication describing the Mach-Zehnder solution to the "sidebands-on-sidebands" problem, and the resulting noise analysis.
• Lab Infrastructure:
  • Steve bought a new maglev turbopump to replace the old and failing one. The new one requires cooling water. Steve is looking into quiet cooling water supplies, such as the system used to cool our MOPA, or the system used to cool the TCS at the sites.

February 9, 2006

• General:
  • Noriyasu Nakagawa from TAMA is visiting for 3 weeks to learn about our digital control system. Please make him welcome!
  • There was a meeting of the 40m Technical Advisory Committee this Thursday 2/9/06. These slides were shown.
  • A draft of the paper on the measurement of the 40m transfer function, with comparison to the theoretical prediction, is being circulated.
• IFO commissioning:
• IFO modeling:
  • Monica is trying to reproduce the 40m optical transfer function in e2e. It looks close; the demod phase needs tweaking.
  • Osamu continues to develop his e2e model of the length control system, making Matt's model more realistic. He's working with Hiro to make modeler_freq function more like a real transfer function measurement.
  • Rob and Alan worked on fits of the 40m transfer function to the theory formula of Buonanno & Chen.
  • Rob is working on developing his numerical simulation of heterodyne readout in the full quantum two-photon formulation.
• DC detection development:
  • Ben has stuffed The DCPD boards. Preliminary testing exposed a few minor things that need tweaking. Other than that, they seem to function fine, but in-depth testing and noise measurements will be made shortly.
  • Ben got the in-vac DCPD electronics tube holder back from the machine shop and it looks great. The photodiode mount is being fabricated currently, and seems to be going well. Ben will deliver the PEEK for the remaining parts, and have them manufactured soon.
• Electronics, controls, computers:
  • Dan and Monica are making progress in understanding the noise budget code, and modifying it to work at the 40m.
  • The vacuum control monitor is somewhat broken; Jay will fix by the end of the week.
  • Bob installed cables from the MCR camera to the video switch. The switch now needs programming.
• Lab Infrastructure:
  • Steve has been following up on various items that came up during last week's safety inspection.
  • Steve is evaluating different kinds of in-vac beam-dumps for the DC readout bamline, including razor blade dumps, the new Japanese "micro-finger" dump, black glass, and Mike Smith's oxidized aluminum. Bob cleaning and baking the razor blade dumps.
  • Steve is planning on building a beam tube for the MCR beamline as it passes through the PSL enclosure.
  • Steve got a new desktop computer.

February 2, 2006

• There will be a meeting of the 40m Technical Advisory Committee this Thursday 2/9/06 ay 8:30 Pacific. All interested parties are welcome to attend; contact ajw@caltech.edu for instructions.
• We are preparing a couple of papers. One, on the measurement of the DARM transfer function in the full dual-recycled Fabry-Perot Michelson configuration (featuring the detuned RSE optical resonance and the optical spring resonance). Another, on the Mach Zehnder solution to the sidebands-on-sidebands problem.
• IFO commissioning:
• IFO modeling:
  • Rob, Alan and Osamu are fitting the dual-recycled transfer function measured last November to the theoretical curve from Buonanno & Chen, and getting excellent agreement. Rob also gets excellent agreement with his Optickle model. We are now working on errors and degeneracies in the parameters.
  • Osamu and Rob are developing methods for measuring the arm cavity pole to aid in getting the model right. They are also thinking about measuring, and developing a detailed model of, all the analolg and digital filtering in the XARM and DARM loops.
  • Monica continues to work on e2e simulation of the DARM open loop transfer function with and without radiation pressure. She's running modeler and waiting for real data to complete the comparison. She's simulating the optical response for different demodulation phases and looking for the values that matches the real measurement. A demodulation phase of pi/15 (12 degrees) could match pretty well : still under investigation.
  • Osamu continues to develop an e2e simulation of the digital suspension system including local damping model with 40m parameters based on Matt's AdLIGO quad model with minimum change. It is almost done and he will move to FP cavity with this suspension model.
  • Monica and Dan continue to work on the NoiseBudget procedure. They are modifying the code obtained from the sites and working on various software problems.
• DC detection development:
  • Rob and Rana continue to develop the detailed schematic of the DC readout controls, electronics, cabling, fast and slow channels, DAQ, etc.
  • Ben reports that the DCPD electronics boards are back from PCB Express. He will stuff them and begin testing. He got some PEEK plastic stock that he will bring to the machine shop for the manufacture of some parts for the DCPD.
  • Dan and Sam continue to work out the bugs in the control of the in-vac pzt steering mirrors.
  • Go and Osamu have prepared a conceptual design schematic for installation of the MIT vacuum squeezer into the 40m IFO.
• Electronics, controls, computers:
  • linux101, the vacuum operator linux workstation in the IFO hall, had a few problems, all of which are related to networking. Rob patched the problem, but it needs the attention of someone who knows how to properly set up networking.
  • Similarly, our new operator workstation, op540m, sits dormant, since we don't know how to get it to network. Waiting for help from Alex.
• Lab Infrastructure:
  • We had a safety inspection last Friday. Steve and Bob escorted the inspectors around the 40m lab and South Annex Lab. A list of action items were compiled by Steve, and all are being followed up on.

January 26, 2006

• IFO commissioning:
  • Rob and Rana made some progress in deterministic locking (lock acquisition by slowly bringing mirrors into alignment without needing to wait for them to swing through resonance in an uncontrolled way; and using normalized DC signals which are insensitive to changes in other degrees of freedom). They can deterministically lock the PRFPMI, with either a CARM offset or a DARM offset. Re-aligning the SRM didn't work, but they got lots of ideas for things to try next.
  • Dan and Monica are making some progress in understanding the noise budget code and adapting it to the 40m.
  • Osamu and Ben are at LHO for shifts. Rana is on vacation. Not much activity this week.
• IFO modeling:
  • Monica continues to work on her e2e simulation of the 40m optical configuration. She and Hiro found a 2pi bug in her pendulum filter; now, the optical spring peak is in the right place. She continues to work on her control system. She is measuring transfer functions with varying demod phases.
• DC detection development:
  • Rob is assembling all the in-vac hardware for the DC readout beamline.
  • Rob is drawing up a complete list of channels, cabling, etc for the DC readout control and monitoring. Looks like we might need another Pentek ADC. And, we need a PZT driver for the output mode cleaner, and PZT drivers for the steering mirrors. The PZT steering mirror on th BS chamber has no available in-vac caling. We might have to drape a cable over from the Input Optic chamber.
  • Dan and Sam continue their measurements of the PZT steering mirror performance. There appear to be so many resonances, as low as 150 Hz, that it may prove impossible to get an alignment servo UGF much higher than ~ 20 Hz. They are experimenting with changing the mirror or the DLC mounts.
  • Steve is purchasing various items needed for the DC readout: in-vac beamsplitter with 3degree wedge; SR 830 digital lockin amplifier, etc.
• Electronics, controls, computers:
• Lab Infrastructure:
  • Bob, Steve and Alan are getting the 40m ready for the safety audit that will take place this Friday the 27th.
  • Bob bought a load of booties and gloves for the 40m and South Annex labs.

January 19, 2006

• Osamu's writeup on the status of AdvLIGO prototyping at the 40m was accepted for publication with no changes.
• IFO commissioning:
  • Rob, Matt and Rana have developed a "deterministic" locking scheme that uses normalized DC signals to lock the power-recycled Faby-Perot Michelson in a slow, controlled, deterministic way, starting will all mirrors but the SRM aligned. The FPMI locks right up, and holds lock robustly while the PRM swings. The PRC loop is turned on and locks immediately. This procedure is very repeatable and easy. Its not yet on full resonance, but they'll do that soon. Their one attempt to blindly swing in the SRM did not work, but there's many things yet to try.
  • Rob passed his PhD candidacy exam, and now can get back to concentrating on deterministic locking, DC readout, and development of the OpTickle modeling tool.
  • Rob has released a new update of the fast LSC code, with further refinements of his flexible signal switching scheme.
  • Monica and Dan continue to work on the calibration of the 40m noise curve, and the noise budget. they are reviewing the noise budget code from the sites and are updating various parameters for the 40m.
• PSL re-commissioning:
• IFO modeling:
  • Osamu, Monica, Matt and Hiro continue to work on various e2e models of AdLIGO / 40m optical configuration and controls. Some models include radiation pressure, multiple pendula, seismic motion and isolation, control plants of varying degrees of sophistication, etc. hey are working towards a model that merges the best features of each.
  • Monica's e2e model of the 40m still has a problem with the x arm not locking well. And it still has the optical spring peak in the wrong place, but it does scale correctly with input power. Under investigation.
• DC detection development:
  • Ben received the beamsplitter mount for the DC photodiode from the machine shop, and it looks good. He plans to submit the rest of the design for fabrication very soon.
  • Ben is reworking the electronics layout for the DC photodiode. He will send the PCB files to PCB Express shortly. In the meantime, he's going to begin to test the existing boards for functionality.
  • Ben, Rob and Rana are working on the detailed design of the DC readout controls, including the in-vac cable plant, DC PD electronics, satellite box, etc.
  • Dan and Sam continue to work on characterizing the PZT steering mirrors for the in-vac DC readout alignment control.
• Electronics, controls, computers:
  • Ben is continuing to experiment with the double-demod RFPDs for our LSC system. He knows component values that work, but some optimization is still in progress. He's tTrying out an active notch for tuning the RFPDs. Will measure the transimpedance.
  • Rana and Alan are searching for a scheme to synthesize a Mach-Zehnder (or two pairs of sidebands with no sidebands-on-sidebands) using one EOM.
  • Dan has finished the construction and testing of a low noise preamp box (design by Rai), gain of 100, input-referenced noise ~ 1.4 nV/rtHz, which will be useful for measuring the coil driver noise.
• Lab Infrastructure:
  • The sump pump under the sink in the entry room failed, stinking up the lab. It's been repaired, but we plant to remove that sink (good spot to put a drill press).
  • Steve is going over the fire alarms, emergency lights, etc, in preparation for a safety inspection next week. He found one emergency light with a dead battery, to be replaced.
  • Bob is inspecting the bake-oven lab in preparation for safety inspection next week.
  • Bob cleaned up the storage cage outside the lab. he's clearing the space right outside the clean room emergency exit. Currently, that emergency exit leads to the storage cage, which is locked on the outside. Bob will come up with a solution.

January 12, 2006

• IFO commissioning:
  • Rob has been thinking about deterministic locking strategies, and also quantum noise in 40m and AdLIGO.
  • Rana wrote and tsted a script to smoothly lock the dual recycled Michelson (DRMI) in a controlled, "deterministic" way, based on an algorithm from Rob (locking the Michelson with misaligned recycling mirrors, then slowly bringing the recycling mirrors into alignment). It works, but will benefit from stronger DC signals, which Rana is working on.
  • Rob has released a new, somewhat cleaned-up version of the LSC code, with gain switching for the DC PD signals, and a "pre-Input Matrix" to implement flexible switching from one LSC error signal to another during lock acquisition.
  • Rana is thinking about how to replace our Mach Zehnder with a single Pockels cell, driven by a signal which minimizes sidebands on sidebands. To that end, he evaluated the performance of the RF module built earlier to generate 4f, 5f, and 6f signals from a 1f input. He found that the module will work well if he adds some RF filters to improve rejection of parasitic frequencies. He has read up on the design of suitable RF filters and says he knows how to do it.
  • Monica and Dan are working together on the calibration of our DARM response and the associated noise spectrum and noise budget.
• PSL re-commissioning:
  • The MOPA laser chiller temperature seems to be increasing; to be investigated.
  • The PSL slow loop was unresponsive, requiring a reboot of the c1psl cpu.
• IFO modeling:
  • Osamu continues to develop his e2e model. right now it's still just a single arm. He has many details implemented (seismic noise, stacks, pendulum, etc), with more to come. He will merge in Monica's detailed optical plant for the full dual-recycled FPMI. The simulation behaves sensibly for lock acquisition of the single arm. He is modelling the common mode rejection of the seismic noise over the 40m arm length: CM noise is suppressed by a factor of more than 10 for frequencies below 10 Hz. Next: more tuning, then move to DRFPMI.
  • Monica continues to develop her e2e model of the full 40m optical configuration, with length control. She can lock the DRMI plus the Y arm, but the X-arm is problematic; needs tuning of demod phases. She measured the OLTF with modeler_fit.
• DC detection development:
  • Rob has started to assemble all of the pieces of the in-vac DC readout beamline on a breadboard.
  • Ben is continuing to work on the Solidworks assembly of the DCPD mount design. I have finished the design for the beamsplitter mount, and have given that to the machine shop for fabrication. The remainder of the pieces should go to the shop shortly.
  • Dan and Sam have set up the DC readout beamline PZT steering mirrors and measured the angular noise in pitch and yaw. They see several resonances; the lowest frequency one is at 120 Hz.
  • Keisuke Goda plans to bring his vacuum squeezer to Caltech in March and begin plans to implement it at the 40m. There is some space that can be easily freed up on the PSL table. We hope to begin the squeezed vacuum experiment by summertime.
• Electronics, controls, computers:
  • Ben has tuned a double-demod RFPD for our LSC system. He has measured the frequency response, and will measure the transimpedance to compare the shot noise with an older LSC RFPD. Now that he knows the right component values, the other two PDs can be stuffed and packaged in the near future.
  • Ben contacted Richard McCarthy for an installation drawing for the new Mode Cleaner servo that we just got. He replied almost immediately with a comprehensive suite of drawings that will make installation as painless as possible.
  • Rana worked on modifications to the the whitening board for the RFPD DC signals into LSC, which will be used for a deterministic locking scheme.
  • Stuart set up a script to incrementally back up all of our minute trends and all of our controls data (scripts, BURT snapshots, etc) to the data achive at CACR.
  • We have a new controls computer (op540m), that needs some set-up (by Alex).
  • Stefan installed rockIFO, a script that pipes any channel you want through any filter you want, to the speakers. Now we can listen to the error signal on whatever loop we're tuning.
• Lab Infrastructure:
  • Bob has been taking care of the lab and vacuum system while Steve is off this week on vacation at Mammoth.

January 5, 2006

• IFO commissioning:
  • Osamu realigned the mode cleaner and re-commissionined the MC WFS system.
  • Rob exercised some configure scripts for locking the interferometer in various configurations, including one and two arms, Fabry-Perot Michelson, and short dual-recycled Michelson. In other words, everything except the full dual-recycled Fabry-Perot Michelson (which comes next).
  • Last night, Rob, Rana, Stefan and Matt got the IFO to fully lock, but it didn't hold long. They switched over to locking with the SRM is the wrong place (opposite side from optical spring) and it seemed to lock reasonably well. Lots of little things to tune up better.
• PSL re-commissioning:
  • Steve measured the throughput of the PSL beam through the PMC (a dismal 58%; we've seen it above 85%) and total throughput from MOPA to periscope is only 42%. Still, it'll do for now.
  • Rana spent some time optimizing the PSL frequency stabilization servo.
  • Steve and Rana redid the layout of the PMC reflected path to give enough light onto the RFPD to get the servo bandwidth up to snuff (about 1.5 kHz). Steve installed a new ccd camera in the PMC reflected beamline and re-installed the SP camera. Bob will wire the new camera into the video switch.
  • Ben and Rana made some modifications to the PMC board, tuning some notches. He also re-set the demod phase. The PMC servo bandwidth is now ~ 8 kHz with plenty of phase margin.
  • Our Mach Zehnder appears to be working well with the new MOPA laser, but throughput appears to be a bit low. We might want to tune up the servo board. But Rana is instead thinking about eliminating the Mach Zehnder, and instead developing a synthesized signal to send into a single broadband Pockels cell which will produce our two pairs of sidebands without generating sidebands on the sidebands (or, only generating very small AM sidebands on sidebands).
  • Rana noted that our mode cleaner servo board is of an undocumented vintage, and he had a hard time figuring out what it was doing. He measured the OLG, and found a UGF of 22 kHz, which is pretty low. He decided to await a new Siggian MC servo board, which LHO will ship to us next week.
• IFO modeling:
  • Monica has used her e2e model to produce a transfer function that shows a clear optical spring peak, although it's at a somewhat higher frequency than expected or seen.
  • Osamu continues to develop his e2e model. He is implementing the 40m seismic noise and stacks, and is developing a detailed model of the controls, including all the analog and digital filtering.
• DC detection development:
  • Mike Gerfen finished machining the output mode cleaner (OMC) body, out of copper. It's beautiful. A few minor fixes to the mirror bezels are in progress.
  • Ben is continuing to work on the Solidworks assembly of the DCPD mount design. He talked to Rick at the machine shop, and he said that the fabrication should go relatively easily.
  • Bob has tested the resistors that Ben and Rana plan on using for the DC photodiode electronics; they are safe for use in high vacuum.
  • Ben plans on testing the DC photodiode circuit that he and Rana developed, in the coming week.
  • Rana and Rob will draw up a detailed scheme for length and alignment control of the DC readout system, and begin ordering parts.
  • Sam and Dan are beginning to build a prototype alignment servo system for the DC readout beamline, using our two in-vac Piezo-Jena steering mirrors as actuators. They assembled the mirrors (without breaking the PZT stacks!!) and wired up one degree of freedom to a spare IO pzt steering mirror board, and verified that it works.
  • Sam is thinking about the alignment servo design. It's not clear whether QPDs or dithering will be used to sense the beam position/angle going into the output mode cleaner. Because any beam jitter going in to the OMC translates into transmitted intensity noise and thus GW signal noise, Sam is aiming to develop a high-bandwidth system.
  • Dan and Mike Gerfen are fixing several small problems with the output mode matching telescope machined parts.
  • Steve will install a "dirty" picomotor into the output mode matching telescope and test the mechanical resonances with his phonograph needle apparatus.
  • The breadboard for pre-alignment of the in-vac DC readout beamline is in the shop, expect it in a week.
• Electronics, controls, computers:
  • Alex Ivanov has succeeded in recovering our 3.7 years of minute trend frames from the failed RAID array, and has made them available once again to dataviewer. Next step is to get a regular and reliable backup system. We hope to be able to push the data to CACR archives in Powell-Booth, with help from Stuart and Dan.
  • Ben has begun the tuning of some double-demod RFPDs for the 40m lab.
  • Larry installed non-ligo-specific GC software onto a new Blade 1500 for use as a new operator workstation (op540m); we'll put it into service by the end of the week.
  • Rana added a bunch of DAQ indicators into frame channels.
  • Rana reworked the DAQ interface board which routes signals from the PSL FSS, PMC and ISS systems to the data acquisition system. He stuffed it with the appropriate amps and filters for all the various signals. The table-top ISS servo board is rev 0.1, needs some mods to get into DAQ. Rana modified the ini files to get the right signals into the frames, correctly labeled.
• Lab Infrastructure:
  • After the rains, Steve reports that there were NO roof leaks. There are, however, plenty of ants.

December 22, 2005

• IFO commissioning, PSL:
  • Rana and Steve have been tuning up the PSL. Rana did a temperature scan on the MOPA, and found a good stable spot far from any mode-hopping. The beamline to the FRC was re-aligned. The PMC transmission is measured to be ~ 76%, which isn't great (we've had more than 85%) but is acceptable. The ISS pickoff beamline was adjusted, and 2 optics were removed from the path.
  • The FSS, PMC, ISS, MZ and MC servos are all working, although their gain and UGF will be optimized once we increase the power.
  • The FSS slow loop was updated to the current version by Peter King.
  • Steve and Rana will redo the PMC reflected beamline to get the right amount of light on the RFPD and the camera.
  • The PSL periscope steering mirror is still unplugged. To be investigated.
  • Still need to adjust the FRC periscope position, and the mode matching.
  • The mode cleaner locks reliably, but the alignment seems off, probably because of a misaligned input beam due to the PSL periscope steering mirror being off. The WFSs don't work in this state.
  • Osamu has realigned the PSL periscope steering mirror and MC. Now, IP-POS and IP-ANG QPDs look good, and the beam is centered on the ITM.
  • Both arms can be locked. The BS needed some adjustment in pitch. The arms have about 0.5 W incident.
• IFO modeling:
  • Osamu continues to work on the e2e model of the 40m, and is now adding realistic seismic motion.
  • Rana has been thinking about lock acquisition with AdvLIGO. Given the available force at the test mass, he estimates that AdLIGO is a factor 50 away from being able to aquire lock. He's thinking about what can be done, especially with suspension point interferometers. He suggests that we use e2e to investigate the lock acquisition difficulties with realistic TM velocities.
• DC detection development:
  • Ben is continuing to work on the Solidworks design of the DCPD mount assembly. Should be finalized in the beginning of the new year.
  • The output mode cleaner is fabricated. Steve is having them build a protective fixture before we bring it to the lab.
  • Lee Cardenas is assembling two DLC PZT mounts for a test of the OMMT steering mirrors. Mike showed Steve and Dan how to assemble the PZT tilting platform to the DLC mounts without torquing the PZTs. Mike submitted DCNs for the changes.
• Electronics, controls, computers:
  • There hasn't been any progress in Alex's attemps to recover the 3.7 years of minute trends on our failed RAID array.
  • Ben is finishing a diagram of the new TT FSS as implemented at the 40m.
  • Five RFPDs are well underway in their final assembly at Wilson House. They will soon be ready for Ben to tune their RF frequencies.
  • Steve notes that there are still OSEM glitches on MC2 UR, although most of the other OSEMs look ok. This is believed to be a problem with the ICS110 ADC in one crate.
  • Mike Pedraza is setting up a new operator consolde (Blade 1500) for us.
• Lab Infrastructure:
  • The portable air conditioner in the control room is bad; out for repair.

December 15, 2005

• IFO commissioning, PSL:
  • Rana and Steve have been working on re-commissioning the PSL. The old FSS board seems to be fried, possibly due to a flaky power supply. So they went ahead and implemented the new tabletop FSS (TTFSS). They spent a long time debugging before they found and replaced a flaky cable. Now the TTFSS is in and working, coarsely tuned, with a ~<1% FSS pickoff. The UGF is 400 kHz with plenty of gain margin, and the gain slider is maxed out. Steve has installed a ~4% pickoff, so we should be able to get the UGF above 650 kHz. Rana tuned a notch on the phase-correcting pockels cell. He also needs to tune a notch on NPRO PZT (fast) path. The FRC periscope is 3-5 mm offset from beam center; will be tweaked.
  • The FSS RFPD has some DC offset on output, and the resonant gain is not in right place. Ben will tune it up with the AM laser setup at Wilson House.
  • The PMC loop gain is low, because the beam on RFPD is attenuated. Rana and Steve will rearrange the PMC reflected beamline to simplify it and get more light onto the RFPD, and Steve put a camera there. (He borrowed it from the SP beamline; now we need a camera there!). Rana's mods to the LLO PMC board are still under discussion at LHO, so for now, we will continue to use our old PMC board.
  • Autolocking scripts for FSS and PMC work.
  • Rana installed and ran a script to do a slow scan of the NPRO temperature, to put the NPRO frequency in a stable place, far from where it likes to mode-hop.
  • Steve installed a new ISS pickoff. Now there will be plenty of light on the ISS photodiodes (when we turn the input power up to 1 watt or more).
  • The PZT steering mirror on the periscope from the SL to the MC has some electronics problems. Rana, Steve and Ben are checking it out.
  • When Rana tried to lock the MC, he noticed that MC2 had anomolously high force-to-pitch coupling. Indeed, Dan found that the SUS Pos gains were all screwed up for unknown reasons. He set them to 1.
  • Osamu and Rana realigned the beam path, starting with the PMC, then the Mach Zehnder, and the periscope, each time gaining significant amounts of transmitted power. Alignment to the MC is pretty good now, and the MC is locked again.
• IFO modeling:
  • Rob is at Hannover for a QND workshop. He's giving talks on the optical spring as observed at the 40m, various modeling tools used to understand the AdvLIGO configuration, and his new modeling tool OpTickle. See QND.
  • Osamu worked on e2e and Bench modeling of the 40m noise. He verified with e2e that the RF demod phase of the photodiode at the AP changes the effective homodyne phase, as predicted by Kentaro, and the transfer functions measured with e2e with different RF demod phases agree well with the prediction from the Buonanno/Chen theory with different homodyne phases. This means that we can investigate different homodyne phases with the 40m, even with RF readout!
  • Osamu notes that e2e is currently the only software tool that has both RF readout and radiation pressure effects. Corbitt's code doesn't have RF, Finesse doesn't have radiation pressure. The OpTickle code being developed by Matt and Rob will have both.
  • Osamu has been modifying Bench to get more realistic predictions for the seismic, thermal and suspension noise. He predicts that it may be possible to observe the dip in the noise spectrum associated with the optical spring if we can get all technical noise sources down.
• DC detection development:
  • Mike has changed the drawings of the OMC beamline components to shorten their heights accordingly to accomodate a baseplate. Updated the 40M optical layout to add the changes. Steve is making the new baseplate. CES is modifying the OMC reflected beam steering mirrors mounts, and is fabricating the OMC. Lee Cardenas is assembling two DLC PZT mounts for a test of the OMMT steering mirrors. Mike will show Steve how to assemble the PZT tilting platform to the DLC mounts. He is in the process of submitting DCNs for the changes.
  • Rana noticed that the in-vac DC readout beamline has no beam dumps. He and Steve will get some black glass.
  • The AP table will need some rearranging for the OMC reflected and transmitted monitor beamlines. Rana will see if he can use Mike's ACad layout.
  • The MIT squeezing group propose to build a 2nd level on our AP table to hold their squeezer. We will look for simpler alternatives that don't require periscopes.
  • We would like to dispense with the Mach Zehnder and drive a single pockels cell with one cleverly synthesized waveform to get all our sidebands and no sidebands-on-sidebands. We'll talk to Reitze about the work they're doing on this at Florida.
  • Dan and Sam are discussing how to proceed with testing and developing suitable drivers for the in-vac PZT mirrors for DC readout.
  • Rana is thinking about how to most easily control the OMC length and input beam alignment, using all-digital dithering and demodulation, with existing hardware (so we wouldn't need to but an analog lock-in).
  • Ben is continuing to work on the DCPD mount design. The PD electronics prototype is assembled and waiting to be tested.
• Electronics, controls, computers:
  • Ben and Rana installed the Table Top FSS following the installation plan E040423-00. Ben has drawn up a schematic for the system, and assigned cable names to the cables.
  • Some of the FSS signals to EPICS were being pulled down by the DAQ interface board. Ben discovered that the board was not wired correctly, and he fixed it. We still need to check out the signals through the Sander box to the DAQ.
  • 5 RFPDs are beginning their final assembly for the 40m lab. Ben will tune them to the appropriate RF frequencies next week.
  • Alex restarted our DAQ frame building and logging to our new 1.5 TB RAID array.
  • Alex tried many ways to recover the 3.7 years of minute trends from our old failed RAID array. There's some partial success, but he's still at it. Stuart Anderson came by to help.
  • Stuart is setting up a path to back up our code and frame data onto disk/tape archive at the CACR SAM-QFS system.
• Lab Infrastructure:
  • We have two new 20" LCD monitors on the operator consoles.
  • The air conditioning thermostat failed. It was replaced and recalibrated.

December 8, 2005

• IFO commissioning, PSL:
  • The "new" (50K hours!) MOPA was putting out ~5.9 W when it was first turned on at the 40m lab, and the beam scan showed a long non-gaussian tail to the west. After some power cycling, the power went down to 5.4 W, and the tail disappeared. The theory is that it was on the wrong mode, and now it's hopped to the right mode.
  • The water cooling on MOPA was found to be leaking a bit, so Steve replaced the plastic elbow joints (in which he found cracks) with brass. It's still sweating; Steve and Dan are keeping an eye on it.
  • The MOPA power is still at ~ 5.4 watts. At present, it's being attenuated to ~1/2 watt right after the beam exits the MOPA, while Steve and Rana re-establish the rest of the PSL.
  • Rana discovered that a jumper in the MOPA, required for SLOW temperature control of the NPRO, was missing. He put one in, and now the SLOW control works.
  • Rana fiddled with the optics in the MOPA box to get a pickoff beam onto the AMPMON photodiode.
  • There are some problems with the EPICS readout of the MOPA monitoring signals. Rana traced it to the signals being pulled down by the DAQ interface board. He's working on it.
  • Rana and Steve moved the FSS pickoff to BEFORE the PMC (it used to be after it). The pickoff is just leakage from a HR mirror. Steve will replace it with a 1-2% pickoff mirror.
  • Rana designed and coarsely tuned the mode matching into the PMC and re-established lock. He went on to put in a new lens to get the FSS pickoff beam through the EOM, and then re-established lock of the frequency reference cavity.
  • The plan now is to re-establish mode cleaner lock, then go back and swap in the new TT FFS. Then tune up the PMC transmission and the FSS path. Steve is installing a camera at the PMC reflected port to help in tuning.
  • Steve and Rana plan to install a bigger pickoff for the ISS photodiodes, and Rana will get a good ISS diagnostic going.
  • Rana plans to make changes to the PMC servo board like the ones he made at LLO. We're also expecting new MC servo and CM servo boards, which we will install in due course. We also need to implement the PSL FSS slowpid servo.
• IFO modeling:
  • Rob continues to work on modeling with OpTickle, Bench, Finesse, Corbitt's code. He's preparing a talk to give at the quantum optics workshop in Germany next week.
  • Osamu has been working with Hiro on e2e simulations of the 40m with radiation pressure. He has new predicted noise curves that agree qualitiatively with theory (theory has quantum correlations, e2e does not).
  • Osamu is investigating modification required to the detector calibration procedure when homodyne detection is used. These mods are also required with RF heterodyne detection with unbalanced sidebands, as predicted by Kentaro.
  • Osamu is refining the calculation of the predicted seismic noise at the 40m.
• DC detection development:
  • Sam Waldman will be testing PZT drivers and servos for the in-vac steering mirrors that steer from the AP to the OMC. He will develop a high-BW (~100 Hz) loop to supress intensity noise at the DC readout port. Once Sam has a design, he'll work with Jay to draw it up and build it.
  • Mike will show Steve, Sam and Dan how to mount the PiezoJena PZT tilting platform to the DLC mounts without torquing them.
  • Mike updated his drawing of the in-vac DC readout beamline. He designed a new baseplate to mount the output sub-assy, consisting of the following components: 2 DLC steering mirrors, OMC, GWD beamsplitter, focus lens, GWD, OMC reflected beam steering mirror. He is changing the drawings of the components to shorten their heights accordingly to accomodate the new baseplate. Lee Cardenas is assembling two DLC PZT mounts for a test of the OMMT steering mirrors.
  • Steve and Mike will assemble the output mode matching telescope in the lab, and begin assembly of the rest of the beamline.
• Electronics, controls:
  • Ben has finished running all of the 10 custom cables for the TTFSS installation. I have drawn up a schematic for the system, and assigned cable names to the cables. I will install the new system today, 12/8 following the installation plan E040423-00.
  • Our old RAID array finally seems to have given up. Alex Ivanov installed a new 1.5 TB raid array, and after great struggle, is getting the 4 years of minute-trend frames off of the old RAID array and on to the new. Thanks, Alex!
  • Alex fixed the epics screens' date and time display, by re-starting the NTP daemon on the appropriate cpu (c1dcuepics) and correcting the time.
  • Rob worked on some scripts for autolocking the FSS and PMC servos, with the help of some triangle-wave scanning scripts.
  • Rob continues to work on rewriting the LSC code; don't expect it before he returns from his trip, next year.
• Lab Infrastructure:
  • Dan got his appendix taken out.
  • Kelvin is our new janitor.

December 1, 2005

• IFO commissioning:
  • The replacement laser from LHO (originally on H2; MOPA SN#102 with 42,943 hrs) arrived last week. Steve and Osamu unpacked it Monday morning, and saw no obvious damage. They carefully noted everything that was different between our old faithful laser ("Assisi", after St. Francis of Assisi) and this one. They placed it on the PSL table with 6 spherical washers underneath.
  • Steve hooked up the water cooling to the master oscillator with needle valve fully open. There was a leak, and Steve got new parts and fixed it. New elbows were added to water in and out of mopa box.
  • The NPRO power was measured to be 0.8W with the Newport power meter. However, 60-80% of the beam was rejected from the last thin-film polarizer and so most of the light didn't make it to the amplifier, even on the first pass. Rana minimized the reflection at this polarizer by rotating the half wave plate by ~60 deegrees, thereby maximizing input into the amplifier. They measured 5.4W with the calorimeter at the exit of the MOPA. Effort to get current reading on our old power supply failed. Laser was turned off and on during this process and it resulted in power to drop to 5W.
  • They then scanned the beam profile, and saw a nasty tail. After Rana tweaked the PBS, the tail is gone and the beam looks nicely gaussian and round.
  • Steve and Osamu measured the beam waist size and position coming out of the MOPA in preparation for mode matching into the phase-correcting pockels cell (which is now mounted outside of the MOPA box, as at the sites).
• IFO modeling:
• DC detection development:
  • The DCPD PCB boards have arrived and been stuffed. Ben will begin testing shortly.
  • Ben ordered several types of resistors for the DCPD and gave a sample to Bob so that he can evaluate their vacuum compatibility.
  • Mike reports that CES is having technical difficulties with the NC mill, and the OMC body completion has been delayed for a week or two.
• Electronics, controls:
  • From Ben: I have finished making all of the 10 custom cables for the TTFSS installation. I'm going to draw up a schematic for the system, and assign cable names for labelling the cables prior to installation. I will then install the new system following the installation plan E040423-00.
• Lab Infrastructure:

November 18, 2005

• IFO commissioning:
  • Our old laser appears to be serving LLO well. They installed it, got up to 8.9 watts, and recovered 11.5 Mpc inspiral range, in less than a day.
  • Doug Cook is preparing the old H2 laser for us as a replacement. It will have the old (noisy) NPRO. He shipped it on Wednesday and we expect it on Friday. We will begin installation on Monday, but vacations and Thanksgiving will interrupt the remainder of the week.
  • Osamu is working on a paper about the seismic noise and its contribution to the 40m noise budget, with David Blair and Ryan Kinney. He is taking more data, and fixing some problems with the measured noise spectra.
  • Steve has been measuring resonant frequencies of various components on the PSL table, and he and Osamu are correlating these with peaks in the PSL frequency noise spectrum as measured by the mode cleaner. A few shaky mounts have been identified. He will also check the in-vac DLC mounts that we have for the DC readout beamline, as well as Mike's new OMMT.
  • Rob is continuing to rework the front end LSC code to make it more robust, flexible, and functional.
  • Dan and Monica continue to work on developing the noise budget for the 40m.
• IFO modeling:
  • Osamu has been discussing the modeling of radiation pressure in e2e, with Hiro. He is trying to understand the approximations made.
  • Monica has validated her e2e package with Hiro's dual recycling summation cavity. She and Hiro found and fixed all the bugs, and she has verified that all the equilibrium fields are correct, as is the DARM transfer function. She will repeat error signal sweeps. She's also using Matt's 5dof control plant, with mods including 40m filters, to simulate full lock. So far, she can hold lock in e2e for 10 msec. Trying to figure out reason for lock loss.
  • Rob is working with Matt on yet another frequency domain IFO model, with full quantum formalism. This one is matlab-based.
• DC detection development:
  • Ben has been working on the DCPD design for the DC readout. He sent out PCB files, and the boards should be here tomorrow.
  • Ben has ordered the parts for the DCPD electronics, including several resistors that will be evaluated for their vacuum compatability.
  • Mike Smith now has all the parts for the DC readout beamline, (including the OMMT mechanical parts and mirrors, the in-vac mirrors, mirror mounts and steering mirror PZTs, etc) except for those being made by CES (the OMC body and a couple of mirror mounts). Mike has pre-assembled the OMMT, and everything seems to fit.
• Electronics, controls:
  • Ben replaced the mixer on the Mach Zehnder Servo board with a lower, level 7 mixer which should work better with the signal levels that we actually have.
  • Steve ordered a new RAID array for our DAQ system, to replace our dying one.
  • Dan continues to work on suspension diagonalization.
• Lab Infrastructure:

November 11, 2005

• IFO commissioning:
  • Mike Zucker called on Friday and said that LLO needed our Lightwave 10 watt MOPA laser, as their laser was dying. On Saturday morning, Osamu and Steve disconnected the laser and packed it up in an old crate we had lying around (the Lightwave box was nowhere to be found). Bob drove the box to Fedex, but they wouldn't take it, so Mike lined up UPS drivers to drive it all the way to LLO. It arrived on Monday morning and was immediately pressed into service. L1 was back to 10.5 Mpc by the end of Monday.
  • Rick Savage offered to send us the old H2 MOPA laser as a replacement. Hopefully they will be able to swap out the noisy NPRO for a newer-style 500 mW model. He and Doug are checking it out and plan to ship it down to us by Thursday of next week, so we should have it the following Mon or Tues. Thanks, Rick!
  • It will take some time to integrate the new laser into our PSL. In our old laser, the phase-correcting pockels cell was internal to the MOPA box. We will now install one external to the MOPA box, as they do at the sites. We will also take this opportunity to install a new pickoff mirror which will send more light to the ISS photodiodes. We are debating whether to install the table-top FSS; probably we will (at long last). We will NOT move the FSS pickoff (it will remain after the PMC but before the Mach-Zehnder). Steve will clean out the cooling water lines. We'll want to do a temperature scan to maximize the power. We'll then do a beam scan, mode match into the PMC, and measure the beam power everywhere.
  • Osamu is working on the seismic noise paper with David Blair, and has begun writing the 40m lock acquisition and optical spring paper.
  • Monica and Dan began to take some single arm transfer functions as a first step towards a calibrated noise spectrum, but their work was interrupted when the laser was shipped out.
  • Rob is making major structural changes to the length sensing code in the front end cpu. He's getting input from Osamu on the changes he'd like to see. He's adding new features such as ramping of gains and other signals. He's including lots of comments and will prepare full documentation.
• IFO modeling:
  • Monica continues to develop and debug her AdLIGO e2e simulation with the dual recycling cavity, with lots of help from Hiro. Looking to get results by end of week, do some validation checks, then plan to look at the optical spring.
• DC detection development:
  • Ben has been designing the in-vac DC PD electronics. A prototype board run will be sent out tomorrow. He's also looking for appropriate DC PDs, working on the mechanical support for the PDs, and the satellite amplifier that will be needed just outside the vacuum chamber.
  • Mike Smith continues to acquire parts for the output mode matching telescope and output mode cleaner. An alignment target was designed and a purchase order let to ASCO.
  • Steve and Liyuan tested the new HR supermirror from Newport. The transmission was 600 ppm, at 0 and 5degrees incidence. now we have two low-transmittance mirrors for our output mode cleaner.
• Electronics, controls:
  • Our framebuilder RAID disk array filled up over the weekend (due to a large number of extra channels that were added to full and trend frames). Alex reconfigured to store only 40 hours of full frames instead of 48, and restarted it.
  • Meanwhile, our RAID disk array is dying, so we got a spare from John Zweizig, and also put in an order for a new array, with 3 times the capacity.
  • Dan is building a Rai Weiss low-noise preamp box to have around the lab.
  • Dan continues to work on suspension diagonalization. Much of this work can proceed without the laser. But the laser is needed for diagonalization of the mode cleaner suspensed optics.
  • Steve is measuring the mechanical resonances of the PMC using his phonograph needle apparatus. Will have some results by next week.
• Lab Infrastructure:
  • Steve transferred the loads of the ITMY and ETMY stacks from STACIS to jackscrews. Go-no go gauge was used. Sensor voltages changed less than 100mV. Now all STACIS isolators are bybassed and seismic stacks are sitting directly on the floor.

November 4, 2005

• Osamu gave a LIGO/Caltech seminar on the Optical spring and optical resonance in the 40m Detuned RSE interferometer, on November 1. G050568: ppt, pdf.
• IFO commissioning:
  • Osamu measured the optical gain of the X arm. The xarm cavity pole appears to be at 2.8 kHz, not the spec of 1.6 kHz. Osamu believes this is due to analog whitening and/or anti-aliasing filters, not the actual optical response. Using this assumption, he extracts the effective response of the analog filters and applies this to get the optical response of the full optical configuration. This agrees very well with the theoretical expectation, lending credence to the assumption.
  • Rob has outlined a variety of alternative procedures for bringing the interferometer from no-control to full-control in a "deterministic" procedure (ie, no waiting for the mirrors to swing into resonance and the servos to "catch"); AKA "no-bang" locking. He tried one of the simplest methods, with some success.
  • Two paths have been forged to acquire lock and control the interferometer, one by Osamu and the other by Rob, Rana, and Matt. All interested parties will work together to forge one "official" length sensing and control front end code, and one set of lock acquisition and control scripts. We also all agreed to make good use of the ilog and the new 40m Wiki.
  • Dan measured the oplev noise spectra with the loops on and off, in an attempt to identify excess noise as the stack frequencies due to oplev folding mirrors on the stack optical tables. There does seem to be a noticable effect, but it is not large; so, for now, we do not plan any major effort to get the folding mirrors off of the stack optical tables.
  • Dan, Monica and Osamu are going through the calibration of the noise spectrum. Osamu notes that the optical response function at the 40m is considerably more complicated than at the sites due to the RSE and optical spring resonances, so the calibration must be done with more care. We will need to carefully measure the analog electronics in front of the A/D and after the D/A.
• IFO modeling:
  • Monica has implemented Hiro's new dual-recycling summation cavity in her e2e model, and is busy debugging it. She will focus her work on the main goal of learning how to extrpolate from 40m to AdLIGO.
  • Hiro has used his new dual-recycling summation cavity to make a prediction for the location of the optical spring at the 40m, with two different power levels. It seems to agree well (semi-quantitatively) with our measured spectrum.
• DC detection development:
  • Steve ordered and received two HR supermirrors from Newport for the output mode cleaner. He and Liyuan measured the mirror transmittance at 0degrees: one was T=0.08%, other was 0.2%. Steve sent te worse one back, with a promise that they'll replace it with a better one.
  • Mike Smith made some corrections to the OMC and OMMT drawing packages that were released to the DCC. Many of the parts for the OMMT have been received. Mike is scribing numbers on the parts and is in the process of pre-assembly.
• Electronics, controls:
  • Dan is going through the scripts used to diagonalize the mode cleaner suspensions, tweaking them to get excitations right.
  • Our RAID disk array for storing frames went belly up on Friday. On Monday, Alex uploaded new firmware to the RaidWeb RAID array and it works now. No frames were logged from Oct 28 15:59 -to- Oct 31 11:59, 814575568 -to- 814823984, and minute trend frames for a small number of channels were lost.
  • Our RAID disk array for storing frames is dying. We are looking into getting a replacement, and also getting a fiber link to the disk/tape archive system at CACR. In the meantime, we need to back up all of our trend frames, BURT snapshots, conlog database, etc, before we lose them.
  • Larry Wallace is getting us a new Sun Blade 1500 to use as a badly-needed fully-functioning operator console.
  • Ben and Rana are discussing the DC photodiode in-vac electronics. A prototype is being prepared. They are looking for suitable vacuum-compatible resistors. He's also working on the mechanical mounts.
  • Ben is also working on getting new RF PDs for us, and also new fast mechanical shutters.
  • A new common mode servo board (Sigg design) is, supposedly, "in the mail" from LHO. We look forward to it!
• Lab Infrastructure:
  • The STACIS seismic pre-isolators are not useful. Several months ago, Steve shorted out the ones supporting the ITMX and ETMX chambers. He has now done the same for ITMY and ETMY chambers.
  • Rob set up a new online 40m Wiki to post things like To-do lists, procedures, etc. This is NOT a replacement for the ilog!!
  • Rob assembled many months of vacuum RGA spectra into a "movie". Steve gives it two thumbs up.
  • A new (spare) Osaka maglev turbo pump has been received. Our main maglev turbo pump is more than 10 years old, and is showing signs of impending death.
  • Steve has a setup for measuring ringdown frequencies, using a phonograph needle.

October 27, 2005

• Osamu completed and submitted the talk he gave at Amaldi 6: "Lock Acquisition Scheme for the Advanced LIGO Optical Configuration, for the proceedings of the 6th Amaldi Conference on Gravitational Waves", Osamu Miyakawa, 9/20/05, pdf.
• Osamu will give a LIGO/Caltech seminar on recent results from the 40 meter, on November 1, 11 am in the SCR.
• IFO commissioning:
  • Osamu realized last week that the reason we aren't seeing the optical spring peak in the DARM transfer function may be because we were locking the signal recycling cavity in the wrong place, on the "other side" of the spring (so that it's an anti-spring). He flipped the sign of the SRC servo loop, and then found that he needed to flip the sign of several more loops before he was able to regain full lock in the AdLIGO configuration. And there, in addition to the clear RSE peak at ~ 4 kHz, was a beautiful optical spring peak at ~ 40 Hz!
  • Rob, independently, arrived at exactly the same set of sign flips to achieve lock in the optical spring configuration.
  • In this correct locking point, the spot at the asymmetric port is now much nicer. The spot is much rounder, and the optical spectrum analyzer shows that the carrier is now nearly dark. It looks like the correct locking point exhibits "mode healing", while the wrong locking point exhibits "mode damaging". This is not yet understood.
  • It appears to be more difficult to lock in the "correct" point, which Osamu theorizes is due to its proximity to other, incorrect and less stable locking points. It is still possible to lock at the wrong point; it is clear from the optical spectrum analyzer at the asymmetric port when this happens, and one has to break lock and try again. We need to develop a more reliable procedure.
  • Rob and Osamu use different procedures, scripts, and make different use of the analog common mode servo during their lock acquisition procedures. They will work together to find a common procedure that makes use of the best benefits of both.
  • Dan and Monica will calibrate the DARM channel and take a calibrated noise spectrum, and then develop the noise model.
• IFO modeling:
  • Rob used FINESSE to model the DARM transfer function at the correct and "wrong side" SRC locking point, and gets excellent semi-quantitative agreement with the measurements. He then modeled the CARM transfer function with various offsets, and again sees excellent agreement with the measurements, including clear optical spring peaks at various CARM offsets. We continue to learn more about the complex optical configuration that we've built.
  • Monica has implemented Hiro's new e2e dual recycling cavity module into her optical model and is testing and debugging it.
  • Monica is gearing up to study the optical spring in her e2e optical model.
• DC detection development:
  • Mike Smith completed and submitted to the DCC a drawing package for the 40m output mode cleaner. The drawings were also submitted to CES for machining. The OMC will be built out of copper, to provide damping of its vibrational modes.
  • Steve ordered and received two broadband HR super-mirrors from Newport. They wioll be tested by Liyuan to see if they are indeed HR for 1064 light at ~ 5degrees incidence.
• Electronics, controls:
  • Dan and Ben installed and commissioned new RevB coil drivers for the three mode cleaner optics, MC1, MC2, and MC3. Ben put in the new 4116, ran a cable, did all the cross connect wiring, changed the database, and fixed the screens. These modules provide separate paths for the bias voltages and the drive voltages to the coils. Now all 10 suspended optics at the 40m use these coil drivers. Dan transferred all the pitch and yaw biases from the drive to the bias paths, re-established the mode cleaner alignment, and made sure the mode cleaner locked happily.
  • Now that we have RevB coil drivers on the MC optics, Dan can proceed to diagonalize the coils and output matrix to decouple POS, PIT, YAW without fear of introducing misalignments.
  • Dan will collect data on the oplev noise to see if there is a significant contribution from the stack motion. If so, we will want to move the in-vac mirrors used in the oplevs so that they are mounted on the chambers rather than on the seismic stacks. Steve continues to develop plans for this. First priority, at the next vent, is to adjust the OSEMS on the BS, PRM and SRM suspensions to minimize bounce mode coupling, and diagonalize these suspensions better.
  • Ben is preparing fast-triggered mechanical shutters for the sites, and will make a couple for us.
• Lab Infrastructure:

October 20, 2005

• To celebrate the full lock of our interferometer in the nominal Advanced LIGO configuration, we had a party at the 40m on Wednesday. There was quite a crowd, including many people who made major contributions to this effort, and a good time was had by all. Thanks for coming to help us celebrate!
• The 40 Meter Technical Advisory Committee report from the meeting on 10/13/05 is here (pdf). Slides presented at the meeting are here (pdf).
• IFO commissioning:
  • As reported last week, Osamu finally brought the 40m into full RSE configuration, with the arms and PRC in full resonance, last Thursday. This was accomplished, despite considerable noise, through careful loop shaping and gain optimization as the CARM signal was transferred from transmitted light DC to the POX RF and the offset was slowly brought to zero.
  • On Monday, Rob and Matt achieved full lock via a different path, using the POX RF signal to control the analog common mode servo. Because of the high bandwidth of the analog servo, they reduced the noise enough to make lock acquisition relatively easy and robust. Longest lock seen: 128 minutes. Generally locks last until they are broken by operator, dumptruck or earthquake. Unfortunately, there have been several local quakes with mag > 4 in the last week.
  • Both Osamu and Rob and Matt took CARM and DARM transfer functions, noise spectra, etc. There is a clear RSE peak in DARM, and also in CARM with an offset. It should be at 4 kHz, but (depending on lock) it can be as low as 3.4 kHz. Rb and Matt tried to adjust offsets to move the peak, with only limited success. Under investigation.
  • For non-zero CARM offset, there is a clear loss of optical gain at frequencies below 50 Hz, which hampered lock acquisition for a long time. It appears to be consistent with an optical spring effect (as predicted quantitatively by Kentaro). We don't see a peak where we expect it (~50 Hz), but that could be due to larger-than-modeled losses.
  • Rob and Matt automated the lock acquisition procedure. One manual step (unplugging a cable on the common mode servo) was eliminated when Ben implemnented a switch on the board. Now, the detector can be brought to full lock via script, quickly and reliably. The only hard part is the pre-alignment. We need a better pre-alignment procedure.
  • The lock slowly drifts away from its correct point, presumably due to alignment drift. We need an in-lock alignment sensing system!
  • The asymmetric port spot looks awful, presumably due to misalignments, mode mismatches, offsets, etc.
  • Rob and Osamu compiled a long list of near-term to do items, including pre-alignment and autoalignment, mode matching, and also a comprehensive approach to setting correct offsets, demod phases, gains etc.
  • Dan is gearing up to work on developing the noise budget model.
• IFO modeling:
  • Kentaro made some plots to illustrate the CARM and transfer functions, exhibiting optical spring peak and low-frequency-falloff, with varying CARM offset. We need more predictions for varying optical loss, and also for DARM.
• DC detection development:
  • Mike's assembly drawing for the OMC is complete. A drawing package is being prepared for release to the DCC.
  • Steve is investigating the costs for machining Mike's output mode cleaner design in copper, which is a low-Q material that is high-vacuum-compatible.
  • We received a pair of REO mirrors from LHO, and Liyuan measured the transmission to be T=1.55%, resulting in an output mode cleaner finesse of around 200. This is good enough!
  • We also received REO HR mirrors from LHO, but they're for 43.5degrees incidence, we need them for near-0degrees. So we're ordering broadband superminrrors from Newport.
• Electronics, controls:
  • Dan and Ben are working on implementing the RevB coil drivers on the mode cleaner suspensions.
  • Dan continues to work on suspension diagonalization and performance evaluation. Wiring is going in, and the 4116 DAC is in place. The EPICS database and screens are in progress.
  • Ben implemented a switch on the Common Mode Servo, so that one of two input signals can be switched off when the other is ramped up. He made some modifications to the board, and installed some cross-connect wiring. He then modified the database, and screen. Rob tested it out in place, and it is working fine.
  • Rana is trying to acquire a new generation (Sigg) Common Mode servo board for us.
• Lab Infrastructure:
  • Steve reports that, at long last, rain gutters and downspouts have been installed on the 40m roof.

October 13, 2005

• There was a meeting of the 40 Meter Technical Advisory Committee (by telecon) on Thursday, 10/13, at 8:30 am Pacific. We reviewed the status of lock acquisition, e2e simulations, and DC readout design and procurement. Slides here (pdf).
• Nergis reviewed a proposal to inject squeezed vacuum into the 40m interferometer, sometime next spring. Slides here (ppt) and proposal here (pdf).
• IFO commissioning:
  • There was steady progress towards full lock of the 40m, with Rana, Rob and Matt (RRM) working at night, Osamu during the day.
  • RRM implemented the analog common mode servo (using the REFL DC signal) to the PSL AOM (the MCL path), with a UGF @ ~3 kHz, (still being tweaked). They got a factor 10 rejection of noise at ~ 200 Hz. The only CARM actuation on the ETMs is a slow script servo to keep the analog MCF error signal close to zero.
  • They then were able to reduce the CARM offset, to the point where the power was ~ 80% of the way to full arm resonance. That brought them "over the hump" of the REFL RF (166I) signal; they saw the signal rise and then fall towards zero as they reduced the CARM offset. So, when they manage to smoothly hand off the signal from REFL DC to REFL RF, we're pretty sure it will bring the ifo into full resonance. But it's tricky to hand off the signals in the analog servo (would be easy with the digital servos). Much of this success was due to careful loop filter shaping.
  • RRM deliberately misaligned the arm cavity so that any beam jitter would couple more strongly to the arm cavity error signal. But they did not see any change in the noise above 50 Hz. They conclude that beam jitter (eg, from noisy in-vac PZT steering mirrors) is not a significant contributor to the noise aboe 50 Hz. The PZT mirrors could still be introducing significant frequency noise, present on the arm cavity error signal but not the mode cleaner error signal (the mirrors are in between).
  • Osamu, meanwhile, was using the POX+POY RF signals to control CARM and reduce the offset. Switching from TRX+TRY DC signals to POX+POY RF is now easy and routine. After much careful loop shaping, he was able to also "get over the hump". Late Thursday night, he at last brought the IFO into full resonance, ie, full AdLIGO configuration. He was able to hold lock for many minutes, and regain lock when lost. Arm light power was ~ 78 (arb units), very close to our expectations given the losses in the arms. He took transfer functions (clear RSE peak, no sign of an optical spring peak) and noise spectra. Hooray!
• IFO modeling:
  • At the 40m TAC meeting, Ken emphasized the importance of reliable e2e modeling to address a variety of important issues in AdLIGO.
• DC detection development:
  • Mike has completed the part drawings for the output mode cleaner, and awaits a decision on the material. We want a well-damped material, so that the PZT driving does not excite the mode cleaner body resonances too much. We were considering brass, but Dennis, Dave Tanner and other TAC members cautioned us against it, due to poor high-vacuum properties. Seems like copper is the best choice, so we are now investigating availabiliy, cost, and machining issues.
  • Rob and Rana are thinking about the design of the DC PD electronics, preparing for potentially large power (up to 250 mW).
  • Rick and Betsy at LHO have sent us two REO supermirrors with T=1.4%, and five HR mirrors (at 43.5degrees incidence, P-pol) for our output mode cleaner. We need the HR mirror to work at ~ 10 degrees, P-pol. Helena and company will test the HR mirrors at different angles; they may work quite well. Along with a curved mirror from Peter King, this means we have all the mirrors we need for the OMC.
• Electronics, controls:
  • The ETMY oplev laser diode died. Dan changed it out and got the oplev servo working again.
  • Dan and Ben are preparing the hardware and software to at last install RevB coil drivers on the three mode cleaner suspensions. These have separate bias and drive paths, so the diagonalization procedure will not throw off the alignment of the mode cleaner.
  • In preparation for moving the MCR RF PD to greatly reduce the beamline length (and therefore the potential for misalignment and clipping), Bob put in new RD PD cables on the AP table.
  • Ben has delivered new RFPDs for the sites, and is now preparing 4 new ones for us.
  • Ben continues to develop the design for his fast meachanical shutters.
• Lab Infrastructure:

October 6, 2005

• Matt Evans is visiting for a couple of weeks, and is working closely with Rob and Rana on locking. Osamu is still at LHO working on commissioning tasks. Ben is off getting married.
• There will be a meeting of the 40 Meter Technical Advisory Committee (by telecon) next Thursday, 10/13, at 8:30 am Pacific. All are welcome; see Alan (ajw@caltech.edu) for details.
• IFO commissioning:
  • Matt wrote a "watch-down" script which watches for lock loss, runs the down script, and then runs the tickle script to reaquire lock.
  • Rob, Matt and Rana have been making good progress towards full lock. As usual, locking all 5 length degrees of freedom is routine; but CARM is offset, so that the arms are not at full resonance and the power recycling isn't kicking in. They have automated the lock acquisition, and the step-by-step reduction in the CARM offset. Until this week, it was possible to climb ~ 20% of the way up to full resonance, but then the noise would overwhelm the servos, the arm power would fluctuate wildly, and lock is lost. This week, after much work, they're ~ 50% of the way up to full resonance, and power fluctuations are much reduced, so they can hold lock in this state for as much as 10 seconds. Frequency noise is the biggest problem.
  • The above progress was made possible by
    • many improvements to the servo filters, permitting increased gain in CARM and DARM.
    • They also improved the MCL path, reducing the crossover frequency and thus the frequency noise.
    • They also diagonalized the short degree of freedom loops; now, there is significantly less noise on these.
    • They also found and fixed bugs in the length sensing code which forms the (non-linear) CARM error signal from the DC transmitted power; but they suspect there may still be a problem there, limiting gain at critical frequencies (e.g., 30Hz).
    • They also looked for saturations of RF demod signals in the LSC rack, but found no smoking guns.
    • They also attempted to figure out why the signal coming from SP 166I (RF signal for CARM) is so small. In the process, they burnt the PD, and then replaced it. Work in progress.
  • They are trying to get the digital common mode servo (to MCL) working again, in progress.
  • They are now at the point where they are unable to increas the CARM gain because of apparent mysteries in the optical plant at lower frequencies. Under intense investigation. It may have something to do with the optical spring; Rana estimates that it should show up somewhere below 40 Hz.
  • Many more ideas and things to try; intense effort underway!
• IFO modeling:
  • Monica continues to make progress in simulating sweeps through arm resonances in the full IFO, at various velocities, and comparing with measurements.
• DC detection development:
  • With help from Peter King, Peter Fritschel, Helena Armandula, and the folks at LHO, we have been promised high-quality optics for our output mode cleaner, without having to order anything new. Many thanks!
  • Mike Smith has redesigned the OMC to reduce the proximity of scattering surfaces within the beam path. In the process of making final part drawings for the OMC. Redesigned the mounting pads to increase the lowest resonant frequency. Analyzed the resonant modes using ALGOR: with "red brass" material, the lowest normal mode frequency is approximately 670 Hz with four mounting pads pinned.
  • Parts (optics and such) are coming in for the DC readout beamline, and Mike is gathering them.
  • Nergis and Keisuke Goda have prepared a preliminary proposal for experimental demonstration of a squeezing-enhanced gravitational wave interferometer at the 40m; see Nergis for details.
• Electronics, controls:
  • At Rana's suggestion, Steve is looking into the design of beam paths for our optical lever beams which eliminate all on-stack mirrors, thus decoupling the oplev measurements from the stack motion (especially the resonances at 3-5 Hz). Steve is formulating a plan, but we do not intend to implement it until we vent for some other reason.
  • Steve has placed our 6 accelerometers at strategic places around the lab and is cabling them up into the PEM system in a semi-permantent way.
  • Steve is looking to procure a fiber-coupled AM laser for in-lab testing of RF photo-diodes.
  • Dan has run the MC diagonalization script and adjusted the output matrices, and has re-adjusted the biases to get the MC aligned the way it was before. In order to avoid mis-alignment during the diagonalization procedure, he (and Ben, when he returns) will implement RevB coil drivers on the MC suspensions in order to decouple the biases from the drive voltages.
  • The oplev laser diode for MC2 died. Steve and Dan replaced it. We need that oplev to adjust coil gains to decouple POS-> ANG.
  • The MC3 UL OSEM sensor is flaky; sometimes it reads zero. Bob replaced the cables from the satellite amp to the vacuum feedthroughs, but it didn't help. Steve swapped the satellite amp itself with a spare. If it reads zero again, Bob will replace the long cable from the satellite amp to the rack (first, just the connector on the satellite amp end).
  • Rana and Rob tried to debug the problems with the end QPD whitening switch controls, but they need Jay's help. It's some problem in the Wiring/code/DB.
  • Rob went through all the EPICS channels that are useful for length sensing (and SUS_OVERFLOW), and added them to the trend data (MANY new channels).
• Lab Infrastructure:

September 29, 2005

• Osamu is at LHO for a couple of weeks to help with pre-S5 commissioning.
• IFO commissioning:
  • Rob, Osamu and Rana have been working on bringing the IFO into full lock with resonant arms. Osamu was able to transfer control of CARM from DC to RF at POX and/or POY pickoff RFPDs, but has not yet succeeded in switching control to SP 166.
  • Rob and Rana tried several things to aid in reducing the CARM offset. These include: smoothing over the transition from high-gain to low-gain in the DC signal from the transmitted PD; changing the 'optical plant compensation' filters in the CARM bank; lowering the DARM loop UGF and changing its filters to avoid instability as the light power increases; checking carefully for coil driver saturations. Most of their changes helped. But, noise continues to drive large power fluctuations and lock loss before full arm resonance is acheived. Noise reduction efforts continue.
  • Rana measured the relative gain of the MCL and MCF components of the MC servo, and the frequency of the length/freq crossover. He measured it to be ~ 90 Hz. He then lowered the crossover to ~60 Hz, near the peak of the phase bubble, which reduced the frequency noise as measured by the XARM error signal in the 60-100 Hz region.
  • Rob and Rana continue to hunt for different lock acquisition strategies, both for the 40m and for AdLIGO.
• IFO modeling:
  • Monica continues to work on locking studies with her 40m/AdLIGO e2e simulation. She is simulating sweeps through arm resonances in the full IFO, at various velocities, and plotting properly normalized signals to look for the largest and most promising ones for controlling CARM.
  • Rob is developing his AdLIGO Finesse modelling tools, and sharing them with Rana and Peter Fritchel.
• DC detection development:
  • Mike Smith is in the process of making final part drawings for the OMC. The Solid Works model was set up parametrically, so the mode cleaner length can be changed and the parts will adjust automatically to accomodate. Ken Mailand is helping him analyze the normal modes of the structure with Algor.
  • Steve has been looking into the availability and pricing of brass for the output mode cleaner, which should have better-damped internal resonances. It is also vacuum compatible, relatively easy to machine, relatively low cost, and relatively low thermal expansion.
  • Steve has spec'ed vacuum feedthroughs for the in-vac DC PD electronics can.
  • Rana suggests building as much of the in-vac DC beamline as possible on a single breadboard, so that it can be pre-aligned in air before installing into the output optic chamber.
  • Rana and Ben have been looking at a different DC PD circuit design to handle hundreds of mW of input light power. They will route the resulting current into a high-power vacuum-compatible resistor, which will produce up to 5 watts of heat, sinked to the table.
• Electronics, controls:
  • Dan has finished the diagonalization of the three mode cleaner suspended optics (M1, MC2, MC3) and has updated the output matrix for all 3 MC optics. He attributes improved noise at the bounce mode to his efforts.
  • Dan is now taking measurements to evaluate the improvement in coupling between POS-PIT-YAW in the MC suspensions. This is important because the MC alignment servo injects frequency noise onto the beam due to ANG->POS couplings, so the gain is turned way down. Next, Dan will see whether the MC alignment servo gain can be turned back up.
  • A limitation on the MC suspension diagonalization is the large bias voltage applied to the suspension PIT and YAW, especially on MC3. Dan and Ben are now getting RevB coil drivers onto the MC suspensions, to eliminate this problem.
  • Rob and Dan found and jiggled a flaky cable on the MC3 satellite amplifier box. They will get Bob to replace the connectors or the whole cable.
  • Rob wrote some scripts for gain matching the transmitted light QPDs and thorlabs PDs, so that the length sensing system can switch between them smoothly.
  • Rana suggests suppressing 60 Hz/harmonics coming in to the LSC demod boards through the RFPD cables, by winding the RF lines around ferrite donuts, and maybe even coupling through 1:1 transformers.
  • Rana suggests adding passive filters to the LSC signal whitening boards to suppress HF noise, as is done at the sites.
• Lab Infrastructure:

September 22, 2005

• IFO commissioning:
  • Osamu has been routinely locking the dual-recycled Michelson (DRMI) with one arm, bringing that arm to full resonance. After more optimization of servos, the RMS noise is 15 times less than it was 2 weeks ago. Then he got to full lock of DRMI+XARM. Then he moved on to DRMI+YARM. At first, the YARM was much noisier, but he traced it to a bogged-down suspension controller; rebooting fixed it. Now, both systems lock easily and have very comparable noise.
  • There's still lots of noise in the 100-200 Hz region, some of which is also seen on the MC servo. We figure that most of it is acoustic noise from the PSL table, or vibrating components between the mode cleaner and the main IFO. It might be vibractions or acoustic noise on the MCR sensing beamline; to be investigated.
  • Now Osamu is working on DRMI+DARM+CARM. As usual, he locks CARM offset from resonance, with the DC transmitted light. We have not succeeded in switching the CARM signal to the RF signal at SP 166; the signal is small and noisy. Osamu is looking for a better, stronger signal.
  • Monica can now successfully interpret FP flashes containing doppler wiggles, in terms of mirror velocities. She can extract the velocities two different ways. Both methods get ~3e-7 m/s, in both arms. E2e simulations reproduce this well. She can also extract the cavity finesse and cavity pole, but it's not clear whether the data are being interpreted correctly. Under study.
  • Dan and Monica continue to learn about the noise budget model.
• IFO modeling:
  • Rana is beginning to explore different control schemes for AdLIGO.
• DC detection development:
  • Mike Smith begun ordering optics and parts for the DC readout beamline.
  • Mike has designed a monolithic output mode cleaner body in SolidWorks, for a 4-mirror cavity. Some views can be seen here and here. Mike advocates making this out of stainless steel to reduce the thermal expansion; but others in the group are not so worried about that, and advocate aluminum or brass because it's much easier and cheaper to machine.
  • Jay says that we will design custom drivers for the PiezoJena PZT steering mirrors which will steer the beam from the IFO to the output mode cleaner.
• Electronics, controls:
  • Dan continues to work on the diagonalization of the mode cleaner mirror suspensions, using ezcalockin. He's already tested the procedure on MC1 and MC2, but MC3 is giving him trouble; under investigation.
  • Osamu discovered that the electronics chain from the POY RFPD was broken. Ben fixed the logic circuit in the box, and Bob replaced the cable. It is now fixed.
  • Ben has finished the cleanup from the RFPD DC signals that now come via the P2 connectors. The cables that were lying on the tables after they were disconnected are now removed from the tables.
  • Ben continues work on his fast shutter prototype.
  • Jay and Alex repaired the linux box that runs the c1dcuepics and c1losepics processes. They swapped out the loaner linux box and put in the repaired one. Everything appears to be operational.
  • Jay and Rolf are thinknig about the changes required to our framebuilder to have frames shipped to QFS storage (as well as logged to a local RAID array, as is done now), which is closer to what the sites now do.
• Lab Infrastructure:
  • Steve ordered a replacement for our dying maglev turbopump.
  • Steve is ordering new, lightweight and very clear laser glasses, OD 5+ for 1 micron light.
  • After this week's rain, Steve had a grounds crew clear off the leaves and cat doo-doo from the roof. He says that PMA promises to install rain gutters on the 40m building in October.
  • The "Fleming Cannon" is being moved to Steele House and will be fired, in close proximity to the 40m lab, around 7 times per year. The Student Affairs office promises to notify us by email 24 hours in advance of each firing.

September 15, 2005

• IFO commissioning:
  • Osamu has been working on locking the dual-recycled Michelson with one arm (DRMI+XARM, YARM blocked), since it should be much easier to lock than DRMI+DARM+CARM during the day.
  • He believes that the difficulties in acquiring lock are due to a considerable amount of excess noise on the XARM error signal (with DRMI locked), at the level of 1e-10 m RMS. After some work, he reduced the RMS noise by factor 10, to ~ 1e-11 m RMS. He estimates that locking the full IFO will require an additional reduction in RMS noise of a factor ~3.
  • There was considerable MC2 suspension noise around 16 Hz; reduced by reducing the MC length gain and applying a resonant gain stage in the XARM loop. Broadband noise and many peaks around 100 Hz also show up in the MCL error signal, probably due to PSL table vibrations; reduced somewhat be modifying the XARM loop filters. Many peaks which don't show up in the MCL error signal, presumably from vibrations of in-vac components in between MC and IFO (MMT, PZT mirrors).
  • After the reduction in XARM noise, Osamu was finally able to lock the XARM with a DC offset, switch to the RF signal from POX, smoothly reduce the XARM offset to zero, and resonate the beam in the XARM with the DRMI locked. The power buildup was in good agreement with what is expected from the model (including the arm losses as measured). The lock was fairly stable and lasted for 11 minutes, despite the trucks going by just outside. Next: try to do the same with the YARM, and then DARM+CARM.
  • Rob tried "kick & pull" locking Tuesday night. First he locks the dual-recycled Michelson (DRMI), then DARM (RF signal, no offset) and CARM (DC signal with offset). Then he turns off (holds) the DC CARM servo and turns on the RF CARM servo, gently kicks the CARM degree of freedom and then pulls it back to slow it down, and watches the signals (hoping that the RF CARM servo will catch and lock). No success, but lots of data to look at.
  • Dan and Monica are working on developing a noise budget for the 40m (building on Ryan's summer work).
• IFO modeling:
  • Rob has employed his Finesse model to investigate what happens to the length control signals when the PRC is undercoupled (as appears to be the case for us). There is no significant qualitative change, but the signals all get smaller (ie, a gain reduction). He estimates that the observed arm reflectivity of 85% (which results in an undercoupled PRC with a gain of 5 instead of the design of 14) is consistent with round-trip losses of 200 ppm. Much larger than we'd hoped for, but not unbelievable.
  • Monica is working on her report for Advanced VIRGO. She'll be in Europe for STAIC and other meetings, in October.
• DC detection development:
  • Following suggestions from the 40m team, Mike Smith has modified the ACad drawing of the in-vac beam paths in the DC readout beamline.
  • Mike has updated the parts list for the DC readout beamline, and is preparing quotes.
  • Alan has submitted a budget and schedule/milestones for the DC readout experiment, to Carol & David.
  • Mike and Helena are exploring options for obtaining quality mirrors for our output mode cleaner.
  • Jay is looking at the signal paths for the output mode cleaner sensing and control, and is beginning make some detailed drawings.
• Electronics, controls:
  • Dan wrote up a how-to document on balancing the coils on our suspended optics (the ones that have oplevs).
  • Dan is working on implementing a script for suspension diagonalization (pos to angle and angle to pos) for our mode cleaner optics, using our MC length and angle sensors. This is important because Osamu has identified significant amounts of frequency noise and beam jitter arising from our mode cleaner suspensions.
  • Ben built an analog divider circuit for the CM servo, stuffed, tested and delivered earlier this week.
  • Ben got all of the RFPD DC signals that are acquired to come via the P2 backplane connectors in the LSC rack, rather than from separate cables from the PD, in an attempt to reduce 60Hz noise-producing ground loops. All the seaparate cables are disconnected and tagged.
  • Ben is making a bunch of ISS PDs, including spares for the 40m. He and Todd are preparing a bunch of RFPDs for the sites and for the 40m.
  • Ben continues to develop a fast beam shutter, for the sites (S5) and for the 40m. His 2nd prototype starts attenuating the beam in ~1.2 msec.
  • Jay is thinking about how to redo the (currently inadequate) power for the LSC rack.
  • Jay and Alex are repairing the CPU that we use for the suspension EPICS monitoring (c1losepics). We currently have a temporary CPU, which is not configured properly. He hopes to get the repaired system in, in the next few days.
• Lab Infrastructure:
  • Steve says that our maglev turbopump will fail at any time (it is 15 years old) and is getting quotes for a replacement.
  • Our laser safety glasses disappear at an alarming rate (please return the one you took!). Steve is investigating inexpensive replacements.

September 8, 2005

• Steve successfully completed the Escape from Alcatraz Triathlon on August 14, placing second in his age group. Congratulations, Steve!!
• Ric brought over Professor Tang, who wants to build a prototype interferometer, similar to the 40m, in China.
• IFO commissioning:
  • Osamu has been working on employing the fast common mode servo for lock acquisition (it was not designed for this!). He tried it but found that, due to the large amount of CARM noise, the servo ran out of dynamic range. The problem is at the VCO used in the FSS servo, which has a filter which limits the dynamic range (presumably for good reasons). This means that, at present, the common mode servo is useful for direct lock acquisition. However, it may still be useful in handing off the lock from the DC offset state to the RF resonant state (which is where we are having problems). But we have to reduce the noise!
  • Osamu has been studying the noise with only the Xarm locked, as a simple model of the CARM noise. He estimated the residual RMS noise of the Xarm using POX signal. It is very high. He estimates that the rms fluctuation of the Xarm displacement is 10% of the Xarm error signal p-p; if it were any larger it would be difficult to lock the X arm. Extrapolating to the power-recycled CARM loop, he estimates that the noise is 30x too high to hold lock stably. This may be the reason why we can't get into full lock in the first place.
  • So what's the noise? There are many contributions, including: large peaks at 60 Hz harmonics; peaks between 16-200 Hz from PSL table (seen on MC error signal, so they're effectively frequency noise); some peaks not seen on MC error signal, maybe due to vibration of components in-vac between the MC and the main IFO (faraday isolator, mode matching telescope, PZT steering mirrors); broad-band sensor dark noise. In addition, the MC error signal spectrum has bursty excursions. Much work will be required to reduce these noise sources to the point where they don't prevent us from locking!
  • Ben will return to looking at grounding and 60hz problems.
  • Rob tried some new LSC code last week to implement a "kick & pull" on the ETMs to gently send the arms from their offset-locked state towards full resonance. It didn't work, and then Rob found the bug. He will try again at next opportunity.
• IFO modeling and DC detection development:
  • Monica has been attempting to measure the mirror velocities when the CARM DC offset signal is HELD. Rob and Osamu have taken "Doppler data" of the transmitted light in this situation, and Monica is fitting the data. She estimates velocities of around 0.35 um/s in the X arm and 0.26 um/s in the Y arm. She's running e2e simulations to confirm this.
  • The report and recommendations from the 7/19/05 review of the DC readout plans is here, T050168-00-R, Ken Strain for the review committee, 9/7/05. The slides that were presented on 7/19/05 are here.
  • Mike is back and has learned SolidWorks; he's practicing by designing our output mode cleaner. We have been going over his optics parts lists and in-vac layout, and are converging on a plan. We will submit a budget and then start ordering parts, hopefully in the next week.
• Electronics, controls:
  • Osamu is trying to get the normalized CARM signal into the analog input of the common mode servo. We normalize the SP 166I signal by the arm transmitted light. Ben has designed an analog divider circuit to implement the normalization. He and Osamu worked up a block diagram, and Ben made the circuit, and sent the board off to PCB Express to be made. It should be back any day, and will be stuffed and delivered by early next week.
  • Ben is developing a second prototype of a fast shutter (made from discarded disk drives), and is testing it.
  • Dan is writing a how-to document for diagonalizing the suspensions at the 40m.
  • Dan got scripts from LLO for auto-diagonalizing the MC suspensions.
• Lab Infrastructure:
  • All AC filters in the 40m hall were changed.
  • Steve cleaned and checked out the vertex crane. It still works.
  • Steve reports that the suspension OSEM glitching still exists. No progress on fully diagnosing and fixing it.
  • Steve noted that over last 6 months, the MCT power has dropped from 1.8 to 1.5 (arb units). Will follow up.
  • In preparation for installing elements of the DC readout system in the output optic chamber (OOC), Steve dug up measured TFs for the OOC stack. Horizontal TF has a +13db peak at 2.5 Hz, vertical has a +12 dB at 9.5 Hz.

September 1, 2005

• It's the end of the summer. We said goodbye to Ryan, Marcus, David, Ju Li, and Seiji. Thanks for visiting, for all your help and good work, and for your comraderie!
• IFO commissioning:
  • After the ETMX OSEMs were adjusted during last week's vent, Dan adjusted the ETMX coil balancing and output matrix f2a filters, and also diagonalized the input matrix (Shihori's method).
  • Osamu now finds that ETMX noise is significantly reduced in pitch, and there's a small improvement in yaw. The X arm control noise used to be much worse than Y arm; after we fixed ITMX it became much better, and comparable to the Y arm. Now, after fixing ETMX, the noise hasn't gotten noticibly better; it's still comparable to the Y arm.
  • Now, MC2 is one of the noisiest suspensions, and Osamu believes it is responsible for a noticable beam jitter in MCT that may be contributing to our lock acquisition difficulties. Dan will work next on balancing and diagonalizing MC2.
  • Osamu has been locking the IFO in various configurations, including PRFPMI. He believes our lock acquisition difficulties stem from a large amount of excess noise, origin unknown, that shows up even in single-arm lock (suggesting frequency noise or electronics). He's now concentrating on noise hunting with single-arm lock, or DRMI+1 arm.
  • Osamu is working on commissioning the fast analog path for the common mode servo. He found several problems with the board; it looks like it was partially modified by someone but not finished. He fixed these problems, and tried it with one arm locked; he saw a significant reduction in noise.
  • Now he's trying to make the common mode servo work during lock acquisition, not just once lock is acquired, which was what it was designed for. He needs to find a way to normalize the analog signal from the IQdemod board to the arm transmitted power; he and Ben are working on an analog divider.
  • Rob wrote some LSC code to implement a "kick and pull back" of the CARM degree of freedom, to try to get it from the DC offset to the RF no-offset state. Will try it out soon.
• IFO modeling and DC detection development:
  • Monica is implementing suspensions in her AdLIGO e2e code.
  • Monica is working on modeling signal sweeps with Malik's parameterization, and estimating mirror velocity.
• Electronics, controls:
  • Osamu encountered problems using the AWG or anything that makes use of the test point manager. Also, the frame builder couldn't see test points. Alex investigated, fixed the frame builder problem, and suggested we reboot control computers to recover the proper operation of the test point manager. This seems to have worked, but the cause of the problem, and the proper way to recover from it, remains a mystery.
  • Ben did a quick survey of the state of the DC readouts from the LSC RFPDs. Currently, there are three DC readouts set up to be drawn from the P2 connector on the back of the LSC Interface boards' Eurocard crates. Two more are needed, and these cables have been made, and await connection. Once these other two cables are connected to the P2 Interface card, all five will replace the signal that is now being taken from the DC OUT connector on the RFPD box; and these cables will be removed. This whole procedure should eliminate a possible source of ground loop 60Hz noise.
  • Ben continues to develop fast mechanical shutter prototypes for the sites and for the 40m. Delay times of order 550 usec. See his report elsewhere in this Weekly.
  • Ben reports that new LSC RFPDs are being stuffed for the sites and for the 40m.
• Lab Infrastructure:
  • Steve has determined that TP1 is overburdened during pumpdowns because it's backing turbopump, TP2, is conduction-limited. He is planning on installing a spare dry scroll pump to back TP1 during pumpdowns.
  • Steve has turned off the cryopump following last week's pumpdown, and is regenerating it.

August 25, 2005

• IFO commissioning:
  • Ryan Kinney's SURF talk, 8/18/05, "Noise Budget Development for the LIGO 40 Meter Prototype", in pdf or ppt. Final paper in progress.
  • Marcus Ng's SURF talk, 8/18/05, "Design of an Output Mode Cleaner to Enable DC Readout in the LIGO 40- Meter Interferometer", in pdf or ppt. Final paper in progress.
  • After Osamu diagnosed the poor damping and strong inter-OSEM coupling in ETMX (UL and LR OSEMs are particularly bad), we agreed that a short vent to adjust ETMX was justified. This should hopefully greatly reduce the noise in the x arm, making lock acquisition easier. We agreed that this vent should be limited to adjusting ETMX only.
  • Osamu measured the ETMX OSEM couplings in vacuum. Steve and Bob vented the 40m on Tuesday morning. By 2pm, the ETMX chamber was opened. Steve, Osamu, Rob, Dan and Monica did the following: transfered the stack load from STACIS to the offloading jack screws, leveled the optical table, adjusted the OSEMS to minimize cross-couplings (rotated the side OSEM by 80 degrees) and sensitivity to bounce and roll modes, took in-air measurements, took many pictures. They sealed up by 5:30. Pumpdown began Wednesday morning at 7:30, and was complete by 2pm, with no significant problems. All optics came back to the same oplev readings, and the beam came back to the same QPD readings. All cavities were locked, everything checks out ok.
  • David Blair re-measured the coherence between the ITMX and ETMX OSEMs. Before the vent, there was little coherence between ITMX and ETMX, in contrast to the Y arm, which has near 100% coherence from 0.15 Hz to more than 3 Hz. After the vent, the X arm looks very similar to the Y arm. Success!
  • Osamu measured the ETMX OSEM. He sees -28db suppression of POS on the side OSEM, uniform response on all the face OSEMs, significantly reduced bounce & roll mode motion, and improved noise spectrum of PIT and YAW. ETMX has gone from being the worst suspension (in terms of couplings) to the 5th best, and completely acceptable.
  • Dan and Osamu are now working to re-balance the ETMX coils, re-diagonalize the POS2ANG matrix, and re-diagonalize the input matrix. We expect significantly reduced noise spectra.
  • Dan is writing up the suspension coil balancing and diagonalization procedure as implemented at the 40m.
  • Osamu and Monica measured the arm reflectivity and the PRC losses a couple of weeks ago. These numbers imply a maximum achievable PRC gain of 5, undercoupled. Rob will check affect on length signals at SP.
  • Ryan has made a great start on the noise budget for the 40m, with lots of help from Rana and Rob. Dan and Monica will take on the calibration and noise budget task and bring it to completion.
  • David Blair and Ryan have completed a draft of a technical report on seismic noise at the 40m, concluding that there is a ~20% difference in seismic velocity between the x and y arms. They also developed the arm OSEM coherence diagnostic, which was very useful in diagnosing the problems with ETMX and the effectiveness of the solution.
  • Rob has been working with Seiji on understanding the length control signals as the CARM offset is reduced. They lock the full IFO with a CARM DC offset (all other degrees of freedom are controlled by RF signals with zero offset). They then HOLD the CARM servo and let the mirrors swing towards resonance; or, they give the mirrors a CARM kick in the right direction. They have the CARM RF signal (SP 166I) switched on, in the hopes that it will catch lock with zero offset, but so far, no success. They are studying the signals (clear Doppler ringing as nicely described in Malik's Applied Optics paper. Monica is working on fitting the signals.
• IFO modeling and DC detection development:
  • Monica has merged her optical plant with Matt's control plant and LA procedure, and is testing the resulting system.
• Electronics, controls:
  • Ben has completed the controls for the common mode servo fast analog path to the laser frequency. All of the wiring, databases, screens, module procurement modification and addressing, and functional testing is complete. Osamu & Rob will commission the system. (The slow p