Lab Humidity: 39-42%
Virginio continued building the Telescope and he's ready to install mirrors (more on this later).
The ISC Telescope Assy is complete and is ready for Lisa & Matt to layout optics on it. Started making an ICS Assembly Load for the Telescope Sub-Assembly, here.
Lisa started cleaning Ag-coated Telescope optics. Since the humidity at the lab is over 20%, we've decided to hold off on installing until we receive desicant pouches from CIT (humidity is detrimental to the silver coating).
Following Mark B. work on decoupling longitudinal top mass excitation to test mass pitch motion, filters have been designed, installed in the TEST top mass filterbank and are ready to be tested.
After getting some good performance last week on BSC6, more aggressive controllers (feedback - level 2 and Level 3) were installed on BSC1. Feedforward and sensor corrections filters were also implemented but some tuning will be necessary to get the best of these approaches.
I have attached calibrated spectra of stage 2 in the X and Z directions when the ISI is damped and controlled (Damping + Isolation feedback + Sensor correction + feedforward). HEPI at BSC1 is still locked and we can notice the resonance of the pier around 10Hz.
It seems that the sensor correction effect is limited in the Z direction (poor tuning or filters not engaged? – A second measurement will be performed tomorrow).
Betsy, Travis, Giles
Good progress had been made on successfully welding 2 of 4 fibers to the ETMx and it's PUM this morning. However, upon completion of the 5th weld we noticed a small feature had appeared behind the top left corner of one of the prisms on the PUM. While observing the feature for another 10 minutes, the feature grew and became an obvious fracture in the glass mass. The fracture started while we were welding in close proximity to the prism. Once we noticed the feature we stopped welding. Note, the prism takes a compressive load when the QUAD is fully suspended, so it is not yet obvious that we should or should not continue to use this mass. We have been in contact with SYS, in an attempt to catch COC before COB today - we plan to talk in the morning with a wider audience about what to do next.
Pictures of the fracture are on Resource Space in a collection with:
https://ligoimages.mit.edu/?r=25246
The h1pemmx front-end is powered down to move network connections and reroute timing. We anticipate that it will be back Friday by 12:00 Noon.
Valved-in YBM turbo backed by leak detector for 4 hours today -> While valved-in, starting signal 9 x 10-7 torr*L/sec, ending signal 9 x 10-7 torr*L/sec -> While valved out < 2 x 10-10 torr*L/sec
Relatively quite day
- PSL down until noon (ask Michael R. for more details)
- Vacuum Cryopump alarm, Kyle mention refilling of one of them
- Work ongoin in End-Y (ask Kiwamu and Alexa for more details)
Have a great evening.
pablo
The laser shut off last night around 6 pm. It looks like this was due to low water flow through laser head 4. PSL_TRIP_1 shows this shutdown. Channel HEAD4FLOW drops to 0.45 lpm and the interlock trips at this point. The time scale for this plot is 20 seconds.
I turned the laser back on but before I could leave the laser room it shut off again for the same reason. PSL_TRIP_2 shows the trip, as the flow through head 4 drops and trips the interlock. I'm not sure what the threshold here is but it must be around 0.45 lpm. I turned the laser back and and increased the flow through the crystal chiller from 17.6 lpm to 18.2 lpm to try and give some more head room for the flow sensor, but it did not have much of an effect. The flow sensors have always had some erratic behavior, as you can see in the 2 week plot attached. LLO found the filters in their laser heads were quite dirty, so maybe we are starting to see the effect of particle buildup.
I have updated the ODC EPICS settings for all SUS, ISI and HPI ODC channels. I have also updated the safe.snap files containing these EPICS records, so that they are correctly set on reboots. The attached files contain the list of changes made to the safe.snap files (nothing listed in the changes_*.txt file under the snap file name means no changes were made), and the modified files have been committed to the SVN.
Whitening filters for the ISC Dual QPD Transimpedance Amplifier were enabled for the following units: D1002481-S1102815 - Corner Station D1002481-S1102811 - EY R1, a 200 ohm resistor, was reinstalled for each channel to enable the whitening filter.
Vincent and Dave.
Vincent made a temporary change to the DAQ fast INI files for h1isibs and h1isiitmy and performed a 'daq reload' on these front ends. We then restarted the DAQ at 11:21 with no problems.
(Corey, Virginio)
Continue to build up the Telescope and In-Vac ISC Table assemblies.
Spent time contemplating how to make TMS Assy Loads in ICS.
Topic of humidity came up in the Weekly TMS Meeting came up yesterday---namely the negative effects of humidity on Optic coatings. We have a Dust Monitor in the Lab, but it isn't "online". (For a "first Humidity data point", had a humidity of 21% mid-afternoon time yesterday) In the future, we'll be receiving desicant packs from CIT and will make "Ameristat bags/covers" for our assys with optics and place desicant within these "bags".
Lisa has completed cleaning of all 2" optics which go on the In-Vac Table. We will gather opto-mechanical hardware for these optics & then transport this to EX.
Continue to add photos of progress here.
Actually, want to nullify that "first data point". I assumed the Dust Monitor was constantly running, but it wasn't. Once Virginio Started it, this is when we saw humidity levels around ~40%. Now it's worth noting we have had some rare summer rainstorms for the last few days.
Attached are plots of dust counts requested from 5 PM June 18 to 5 PM June 19.
Attached are plots of dust counts requested from 5 PM June 17 to 5 PM June 18.
Following Ryan's recent odc label update, the "damp" ODC bit has been updated for MC1 MC2 MC3 PRM PR2 PR3 ETMY ITMY BS SRM SR2 SR3, and is now properly working (cf exemple on picture attached)
In order to do that, I set every suspensions in their "damped state", and ran the attached bash script to update the "STATE_GOOD" epic value "H1:SUS-${OPTIC}_${MASS}_DAMP_${DOF}_STATE_GOOD", with the "STATE_NOW" epic value "H1:SUS-${OPTIC}_${MASS}_DAMP_${DOF}_STATE_NOW".
The "damp" bit (here, BIT 6 on the picture) is ONLY a function of the filters engaged in the damping filterbank. It won't turn red when a gain is changed for instance.
I updated the safe_snapsots with the recently modified STATE_GOOD epic values
attached are the bash scripts used to update the safe.snap files, and the corresponding changes.
Gerardo and Gerardo Jr. have launched the processing of another PUM - today, a fresh ear was sucessfully silicate bonded to PUM labeled ITM06.
Note the PUMS are labeled as ETMs or ITMs, but functionally they can serve as PUMs for either type of QUAD suspensions.
Mark B. and Gerardo
(This alog is based on final numbers from Gerardo - a typo was discovered in one of the ear positions used in discussions yesterday.)
The PUM "ETM04" had been selected for use in the ETMx quad at LHO. While reviewing its history in preparation for welding, Betsy, Gerardo and Angus discovered that one ear, on the S4 surface, had been flagged as 0.18 mm from the front-back centerline, outside the 0.1 mm tolerance. (It was one of the first to have ears bonded, before the procedure had been entirely perfected.) Even a small front-back error produces a large pitch torque from the weight of the test mass below, which can produce a significant static pitch in the final suspended chain.
The positions of the ears and prisms are as in the first attachment. The prism on the S4 surface is at -0.07 mm, which adds to the pitch torque. The ear and and prism on the S3 surface are at 0.04 mm, which subtracts a small net torque. The net expected pitch imbalance is per the second attachment: 4.3, 5.4, 3.2 and 2.3 mrad at the various masses, from the bottom up.
The average misposition of the ears is (0.18+0.04)/2 = +0.11 mm. Since it is not possible to remove and reposition the ears after several years of curing of the bonds, to salvage the mass requires moving the prisms to an average position of +0.055 mm (the weighting of the prisms is -2 that of the ears because the tension on the associated wires is double, and upwards rather than downwards). Since removing prisms is time-consuming and slightly risky, it is preferable to move just one if possible. The easiest possibility is to move the S4 prism from -0.07 mm to +0.07 mm, so that the average is (0.07+0.04)/2 = +0.055 mm as desired. This means that the PUM will sit at a yaw angle set by the prisms of (0.07-0.04)/326.5 = 0.09 mrad, and the test mass will be a yaw angle set by the S4 ear and prism of (0.18-0.07)/326.5 = 0.34 mrad. These are small enough to be easily trimmed statically and should have a negligible effect on the dynamics. (A comparison of TFs with and without the misposition and fix will be posted as a comment.)
However this moving of the prism can't be done in time to use the assembled welding crew in the next few days, so PUM "ETM03" has been prepared and allocated to LHO ETMx (per alog 6800), and, assuming a quick and successful rework, "ETM04" will be used next up as ETMx at LLO in a few weeks.
Note - the discovery of the out-of-tolerance ear is due entirely to the diligence of Gerardo! We would have scratched our heads with a horribly pitched QUAD in many weeks time if he had not spotted this.
Gerardo and Gerardo Jr. have placed the ETM04 S4-side down in a bath of acetone for the night. Soaking started around 4pm today.
As promised above, I took the production case (20120601TMproductionTM) of the quad model, and created a custom version of the model with provision for perturbations to the ear and prism positions at the PUM, and a custom case (20120601TMproductionTM_PUMpert) with numerical values corresponding to the proposed rework of PUMETM04. These can be found in the SUS SVN at
^/trunk/Common/MathematicaModels/QuadLite2Lateral/mark.barton/20120601TMproductionTM_PUMpert/
I then calculated the equilibrium position and generated top mass to optic TFs for all DOFs. See attached PDF.
As might be hoped, the equilibrium position has essentially no pitch, and a trivial amount of yaw that agrees fairly well with the back-of-the envelope calculation above.
The TFs are not distiguishable by eye from those with the stock model.
New code was loaded to add an integrator to the low noise VCO which allows the frequency to be be stabilized to a set value. The DC offset slider is used as the actuator. Since the frequency is only read once every second, this is only useful for drift control. A temporary Agilent frequency counter has been set up in the CER to independently monitor the VCO frequency (ALS DIFF). At first glance the measurements are comparable.
Drift of free running VCO over 15 hours.