formatted dust trend results to group each monitor

I have modified the check_dust_monitors_are_working script to group the 0.3µm and 0.5µm lines together to make the output easier to read. In clean conditions the 0.5µm count is usually zero whereas the 0.3µm count is usually non-zero.

H1:PEM-CS_DUST_LVEA2_300NM_PCF           Channel Statistics: mean    4.92 max   20.00 std-dev    7.17 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA2_500NM_PCF           Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_LVEA3_300NM_PCF           Channel Statistics: mean    4.90 max   30.00 std-dev    6.71 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA3_500NM_PCF           Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_LVEA4_300NM_PCF           Channel Statistics: mean    4.53 max   30.00 std-dev    7.15 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA4_500NM_PCF           Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_LVEA5_300NM_PCF           Channel Statistics: mean    4.94 max   20.00 std-dev    6.71 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA5_500NM_PCF           Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_LVEA6_300NM_PCF           Channel Statistics: mean    4.44 max   20.00 std-dev    6.43 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA6_500NM_PCF           Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_LVEA10_300NM_PCF          Channel Statistics: mean    3.43 max   20.00 std-dev    5.70 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA10_500NM_PCF          Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_LVEA30_300NM_PCF          Channel Statistics: mean    4.86 max   20.00 std-dev    6.71 (12 hour(s) minute trends)
H1:PEM-CS_DUST_LVEA30_500NM_PCF          Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_PSL101_300NM_PCF          Channel Statistics: mean    0.61 max    3.00 std-dev    0.84 (12 hour(s) minute trends)
H1:PEM-CS_DUST_PSL101_500NM_PCF          Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-CS_DUST_PSL102_300NM_PCF          Channel Statistics: mean    3.97 max   30.00 std-dev    6.88 (12 hour(s) minute trends)
H1:PEM-CS_DUST_PSL102_500NM_PCF          Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-EX_DUST_VEA1_300NM_PCF            Channel Statistics: mean    3.89 max   20.00 std-dev    6.09 (12 hour(s) minute trends)
H1:PEM-EX_DUST_VEA1_500NM_PCF            Channel Statistics: mean    0.00 max    0.00 std-dev    0.00 (12 hour(s) minute trends)
 
H1:PEM-EY_DUST_VEA1_300NM_PCF            Channel Statistics: mean    5.03 max   20.00 std-dev    6.45 (12 hour(s) minute trends)
H1:PEM-EY_DUST_VEA1_500NM_PCF            Channel Statistics: mean    0.17 max   10.00 std-dev    1.28 (12 hour(s) minute trends)

 

Removal and replacement of PSL ISS 2nd loop PD array; installation of strain relief

RickS, JiianLiu, KentaroModushi, HughR

To enable access to bolt holes in the HAM2 table top for ISI damper installation, earlier this week we removed the improvised strain relief for the ISS PD array cables and moved the ISS PD array.   After completion of the ISI damper installation, we re-positioned the ISS PD array (against stops that had been installed before removing it) and installed the strain relief hardware that had been designed since the first installation of the ISS array for H1 (see D1101059-v5.PDF, sheet 2, zone D3).

Photos with identifying titles are attached below. 

Note that photos of the dust monitor screen showing counts measured just inside the West door opening of HAM2 are included.

 

HAM 5,6 AIPs OFF

All bolts except four were removed from HAM 5 door(s) to prep for Monday morning removal. Turned OFF HAM 5,6 AIP power supplies.

2nd purge air port connected

Today we connected the purge air point of use at IMC tube to increase clean air flow, in addition to OMC valve, after receiving reports that clearnoom air flow was overpowering purge air positive pressure. We talked about dialing down fan motors on cleanroom filters. 

The FFUs on the biergarten clean room were dialed down to ~ 50% air flow on Friday.

WHAM ISI DAMPING Installation Progress

WHAM2--Done: All Damping elements installed, TMDs retuned & Installed, all walls replaced--Thank you Sheila & Jeff!

WHAM3--Maybe 85% done, thanks Corey!: Need to install the Corner3 Horz GS13 Can damper on the north side(+X), replace the Corner3 Spring & Flexure access walls.  The TMDs need retuning but the Blade Springs have all been B&K'd.  The TMDs can be installed after the walls are installed.

Ops Shift Summary

TITLE: 09/22 Day Shift: 15:00-23:00 UTC (08:00-16:00 PST), all times posted in UTC
STATE of H1: Planned Engineering
LOG:

14:51 UTC Karen to LVEA
15:40 UTC Corey to LVEA to move the baffle removed from HAM3
15:44 UTC Gerardo to LVEA to turn on purge air connection at HAM3
15:51 UTC Corey back
16:00 UTC Gerardo back
16:37 UTC Hugh to HAM3 to work on ISI damping with Jim
16:50 UTC Jeff B. to LVEA to survey supplies
17:32 UTC Betsy to LVEA looking for parts
17:48 UTC Dave to CER, WP 7162
18:23 UTC Betsy back
18:34 UTC Untripped HAM4,5 ISI WDs
18:34 UTC Dave back
19:10 UTC Robert to BSC chambers for visual inspection
19:20 UTC Dave powering up ASC
19:38 UTC DAQ restarted
20:00 UTC SUS BS WD tripped. Reset.
20:08 UTC Travis to BSC area to drop off equipment.
20:12 UTC Jeff K., Hugh, Jim to HAM3 for B&K measurments. Jeff K. then to BSC for violin mode measurements.
Peter to PSL enclosure
20:39 UTC Jason to optics lab to retrieve parts
20:45 UTC Robert back
20:52 UTC HAM5 ISI WD tripped. Reset. Tripped again. Left tripped.
21:29 UTC Peter back
21:47 UTC Travis to bier garten to help with violin mode measurements
Delivery from RDL for Gerardo through gate. Kyle received.
22:07 UTC Cheryl to optics lab

Jeff and Jim back, Robert to BSC chambers

23:27 UTC Reset ITMX SUS WD and rerequested ALIGNED. Reset HAM5 ISI WD.

Daily Vent Status

HAM2 (Apollo)
- Work Platforms installed around HAM2

HAM3 (Kissel, Radkins, Warner)
- Finished installing all HAM3 ISI damping mechanisms but Corner 3 Horz. GS13 Can Damper. Corner 1 & 2 walls are back on and torqued
- TMDs still need to be retuned and re-installed
- All three corners' have been B&K hammered

BSC3 / ITMX (Kissel, Sadecki)
- Finished -X/+Y and +X/-Y fiber characterization w/ speaker array this time (and measured up 6th harmonic at 3k)
- Set characterization setup to read out remaining -X/-Y

BSC2 (Schofield)
- Characterized resonances of TCS / Oplev steering mirrors

Planned and unplanned damping of MCA2 and MCA1

After Corey and I removed the center of MCA1 (Swiss Cheese), we installed the damping corks on the eye baffle, MCA2 (see figures).

While adding corks, I noticed that the panels on MCA2 did not fit perfectly and bulged out from the frame because the distance between fastening points was longer for the panels than the frames (see figures). This resulted in long stretches of loose sheet metal that vibrated at low frequencies (order of 10 Hz) with large amplitudes and Qs. Hugh had some Class A 1/16” x 2” x 2” Viton pieces associated with ISI table damping that I wedged in between the panels and the frame, giving some damping (see figures). Of course, the best solution would be to increase the size of the holes in the panel so that the fastening would be more forgiving and to use damping washers.

I found a few bulging panels on MCA1 and damped those as well (see figures).

The bulging surface changes the angle and can increase the reflectivity of the panels from the point of view of beam spots (see figures). I checked that the damping pieces wedged between the panel and the frame did not increase the bulges to the degree that it was evident as increased glints in photos from the point of view of beam spots (see figures).

This damping scheme should, of course, be reviewed and, if accepted, we should plan on doing this on all 8 similar baffles at both sites. The only other issue I can think of is melting of the small exposed region from wildly errant beams. However, I think that only rapidly moving wide-angle beams from HAM2 could reach the Viton, and I think we have other Viton on the tables that is more likely to be a problem.

Finally, the central panel of MCA2 was loose where a bolt had apparently broken off during installation, and we added a small chunk of Viton to damp the vibration (see figures).

Robert, Corey, Hugh, Jim

SQZ: ADC and 16bit DAC card removed from h1asc0, models changed

WP7162 Squeezer Install, ASC H/W and S/W changes

Daniel, Dave:

The 8th ADC and the 2nd DAC cards were removed from h1asc0's IO Chassis. Serial numbers are shown in the table below.

The CDS system was configured to no longer run the models h1sushtts and h1susauxasc0 on h1asc0, and to remove it from the DAQ.

The h1asc and h1ascimc models were modified to use the 1st DAC instead of the no-longer-present 2nd DAC. The details of this can be found in the DCC document:

https://dcc.ligo.org/LIGO-T1100472 page 9.

Because of the slight-but-not-zero chance that powering h1asc0 back on could Dolphin glitch the corner station SUS and SEI, we coordinated with Team-Vent for the optimum time to power h1asc0 back on. Jeff K gave permission for the power up and DAQ restart during the lunch hour. Consequently I powered up h1asc0 with no problems. Before restarting the DAQ I took the opportunity to modify h1pemcs to add filter modules in the input legs of the MAG_EBAY_LSCRACK Quadrature Sum circuit so Robert can perform AC coupling tests on the quad-sum signal. The DAQ restart went well. When I noticed that the EDCU was still RED, I found that h1ecatx1 had frozen at 06:30 this morning (Patrick restarted it).

card	serial number
16bit DAC 16AO16	110419-25
16bit DAC interface card	S1102283
16bit ADC 16AI64	110506-77
16bit ADC interface card	S1102495

 

h1ecatx1 EPICS IOC crash

Occurred at ~6:31 AM PDT. Error in attached screenshot. FRS tickets already exist for the same error at both end X and end Y. Restarted.

New Glove Dispenser in LVEA

   The new gloves and dispensers have been setup for the various cleanrooms in the LVEA. The LVEA garbing room (entry to LVEA) has been mostly reorganized. There is still a little to do; please patient while the remodeling is underway. 

   The new gloves will roll out to the reset of the labs around the site and to both end stations over the next few weeks. 

   Your comments and suggestions are always greatly appreciated.       

LVEA Particle Count Trends - 24 Hour

   Posted below is the trend data from the dust monitors in the LVEA for the past 24 hours. 

   The spikes at HAM2 are associated with work on the ISS second loop array, and are under alarm limits. The HAM3 spikes are associated with the crew opening door covers and getting prepared for ISI work. The high counts lay down quickly and remain low while they are working in the cleanroom. I am uncertain what specificity caused the spikes in the Biergarten. These are not a problem as they also lay down quickly to very good levels. 

   The cleaning crew and cleanrooms are doing a great job keeping contamination to a minimum.   

ISI CPS Noise Spectra Check (FAMIS #6916)

   FAMIS Task #6916 - Posted below are the plots for the ISI CPS Noise Spectra. 

DBB scan

Without touching the alignment in to the diagnostic breadboard, a scan of the front end laser was done;
partly to see what the status of the diagnostic breadboard is after being dormant for such a long time.

    The mode scan implies that the output of the front end has 61% in higher order modes.  Something we
know not to be true.  So clearly some alignment work into the diagnostic breadboard is required (this
is not a surprise).

    It should also be noted that whilst the scan was conducted the PMC PZT monitor flashed red.  But the
take away news is that the diagnostic breadboard still works but needs a service call.

Center of Swiss Cheese baffle removed

Corey and I removed the center of the Swiss Cheese baffle today. We also got the corks in the eye baffle but we found a problem and did some extra damping that I will report on later.

ESD actuation strength measurements

Message: Our current charge measurement (Veff) doesn't tell us everything we want to know about how much charge is on the optics, or do a good job of predicting noise in DARM caused by electrostatics.  Here is a way of measuring the ESD actuation strength that could give us a slightly more complete picture of what is happening.  These measurements still don't give us a conclusive charge measurement that we can rely on to understand the DARM noise. 

Motivation:

In an effort to track the charging of our optics, we have been measuring and tracking the bias voltage at which our ESD actuation strength is minimized for a few years (38604) , but we have some clues that this measurement gives us incomplete information about the charge that creates noise in DARM.  After the July EQ, we saw a large increase in the noise in DARM.   There was also a change in the charge measurements, but not a significant increase in Veff (38401).  However, discharging the end test masses partially reduced the noise in DARM, and reduced Veff.  We also saw coupling from the ETMX and ITMY ISIs to DARM, which made noise that was well below the limiting DARM noise.  Discharging fixed this coupling for ETMX.  

New measurements:

Over the last few weeks I've been thinking about ways we could measure other parameters that depend on the test mass charge. Four coefficients (alpha, beta, beta2, and gamma) that describe the ESD actuation, their measurements and calibration are described in T1700446. Scripts for making these measurements and the log files for the measurements plotted here are in the userapps repository, at /userapps/sus/common/scripts/quad/InLockChargeMeasurements (revision 16175).  These measurements take about 10 minutes to complete for all 4 test masses when we are in low noise.  I suggest that after this vent we repeat this measurement in a way similar to the optical lever charge measurements for tracking the charges on the optics. 

Calibration:

The attached screenshots show the results of these measurements of the 4 coefficients, with error bars based on statistical uncertainty (the coherence of DARM with the excitation).  I used two different methods to calibrate the measurements, first calibrating the DAC outputs then using the monitor inputs (where the correct monitors are in the frames, FRS ticket 8995 is in part going to make sure that all the correct monitors are stored).  The applied bias voltage I end up with is about 13% lower if I use the DAC outputs (380V) compared to the result I get using the monitor channels (430V), which results in about a 22% discrepancy in the values of alpha and gamma (coefficients that describe the quadratic ESD response).  I am using the ESD monitors, since they are consistent with the high voltage that is supplied, and the calibration of the monitors is simpler (especially for the ETM HV divers where the schematics are hard to follow).  

Results:

For the ETMs, alpha is around 2e-10 N/V^2, while according to G1100968 it should be 4e-10N/V^2.  For ETMY the measurements are inconsistent; alpha and gamma should be unchanged by the discharging. I looked at the L2 L witness sensors to see if differences in the distance between the reaction mass and test mass could explain these discrepancies, it doesn't.  For the ITMs, alpha is around 1.2e-11N/V^2; a little higher than the value of 7.5e-12 N/V^2 in G1100968.  

The gamma term describes the dielectric force caused by fields which terminate on the conductors surrounding the optic (the cage).  This should also be independent of charge, and I'm not aware of any FEM models that predict the size of this term.  The coefficients are similar for the ITMs (3e-11N/V^2 for ITMY, 2.7e-11 N/V^2 for ITMX) and ETMs (5.4e-11N/V^2 for ETMX, varying between 4-5e-11 N/V^2 for ETMY).  This means that for the ITMs, the forces caused by the field lines that terminate on the suspension cage actually dominate the overall actuation strength of the ESD, not the field lines which begin and terminate on the ESD electrodes. 

The beta and beta2 terms are due to the interaction of two different distributions of charge on the optic with the fields from the ESD electrodes.  The TMDS at EX was used at days 5+6 on this plot (3 cycles) while it was used for one cycle at EY on day 11.  It is surprising that the betas at EX seem to have increased after the discharge, there is only one measurement from before the discharging.  I looked back at the data and indeed, the response to driving the electrodes without any bias applied was smaller before the discharging than after.  It should be possible to estimate charge densities based on these betas.  

Comparing these measurements to Veff:

In terms of the coefficients measured here, the usual charge measurement measures the bias at which the signal response is minimized, V_eff = (beta-beta2)/2*(alpha-gamma).  The fifth attachment here shows Veff calculated from the measured coefficients, this matches pretty well with the results I got from making a longitudinal measurement of Veff in alog 38608 (In that alog I was reporting bias voltages based on the DAC outputs, so they should be about 13% smaller than what is here, which is based on ESDAMON.  These ETMY numbers are rather different from the Veff measured using the optical lever in alog 38604

A new Veff2:

If instead of driving the signal electrode and finding the bias voltage that minimizes the response, we can do the same measurement by driving the bias electrode and finding the bias voltage that minimizes that response.  This is just a different combination of the ESD coefficients, Veff2 = (beta+beta2)/2(alpha+gamma), which is shown in the 6th attachment to this alog. This is also in reasonable agreement with the measurement I made off this Veff in 38608 .  This Veff could also be measured with the optical levers.  

A somewhat nicer measurement of all 4 of these paramters could be made by driving the signal and bias electrodes in common and differential.  We would need to account for the different low pass filters in the analog paths to these electrodes to do the measurement this way.

 

Shiela, Thank you and Jeff for making these coupling measurements. I have been worrying about this coupling
for a long time. It is one of the reasons we made an injected ion system for each test mass chamber so that
one could imagine a regular rate of discharge (monthly?) once we baffled/moved the ion pumps. The discharge is a
temporary fix. As we improve the detector sensitivity we are going to hit a limit from the charge on the 
test mass just from charge hopping. At some point we will simply have to bite the bullet and coat the entire
mass with a mildly conducting (Gohms/sq) coating and "ground" the test mass through a coating on the silica fibers. I thought
the Stanford group is working on this coating.
RW 

HEPI Pump Pressure 45 Day Trends

See attached screenshot.  Things I noted were:

• There is a glitch in both the EX and EY pump pressure on August 22 which likely corresponds to a HEPI trip described in alog 38313. 
• There is a large spike starting around August 12 that trends back to ~nominal which is likely due to HEPI being locked (aLog 38618).  The signal flatlines thereafter. Was the CS pump station turned off after all HEPIs were locked??  I failed to locate an aLog if this is the case.
• There is another glitch in both the EX and EY pump pressures Sept. 17, which is likely due to Hugh restarting the HEPI pump (?, no aLog yet although it was discussed in the morning meeting) after the Sept. 16 power outage (aLog 38675).