Attached is a 5-day trend of the PSL laser head flow rates. All is unchanged from the last report of these flow rates; head 2 is still ragged as usual.
This morning I completed the weekly PSL FAMIS tasks.
HPO Pump Diode Current Adjust (FAMIS 8433)
With the ISS OFF, I adjusted the pump diode currents for the HPO. All currents were raised by 0.2 A, changes are summarized in the table below. I have also attached a screenshot of the PSL Beckhoff main screen for future reference.
| Operating Current (A) | ||
| Old | New | |
| DB1 | 49.1 | 49.3 |
| DB2 | 52.1 | 52.3 |
| DB3 | 52.1 | 52.3 |
| DB4 | 52.1 | 52.3 |
I did not adjust the DB operating temps. The ISS is now back ON, and the HPO is outputting ~154.9 W. This completes FAMIS 8433.
PSL Power Watchdog Reset (FAMIS 3661)
I reset both PSL power watchdogs at 16:51 UTC (9:51 PDT). This completes FAMIS 3661.
J. Kissel Today's measurement reminds us that we always need several weeks of data to really make a statement about trends when it comes to effective bias voltage. See attached data. ETMX is still OK. Today's data from ETMY shows that the effective bias voltage trend we thought we say last week (back towards zero) has been washed away. We see that the effective bias voltage remains as high as ~90-100 [V] in several quadrants. We're still less that 10% actuation strength change in longitudinal however, even after the reaction chain move that decreased the actuation strength further by a few percent (see LHO aLOG 37854). I still wouldn't raise the red flags just yet, but I'm glad that the vacuum team is "spinning up" the test mass discharge system.
I happened to open up the laser's Beckhoff screen this morning and noticed that the flow rate in the power meter circuit
was fluctuating between 1.2 lpm and 0.5 lpm. Apparently it's been doing this for a good fraction of the past week. For
the time being I've set the trip point to be 0.35 lpm - down from 0.4 lpm.
Not certain what the problem is right now. Possibly a partial blockage in an Ophir power meter similar to the
problem experienced over a year ago.
TITLE: 08/01 Owl Shift: 07:00-15:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Observing at 54Mpc
INCOMING OPERATOR: Jeff
SHIFT SUMMARY: locked all shift, out of Observe briefly for violins
LOG:
Attached is a comparrison of how the OSEMs for ETMX, ETMY, ITMX, and ITMY changed from before the Montana EQ to after.
The OSEM sets (for example, pitch, all stages) are color coded:
Just looking at color, ETMX and ITMY are all yellow or red (more likely to have had issues as a result of the EQ?), and only ETMY and ITMX are green and yellow.
Attached is a plot of this lock showing violin modes ringing up starting about 2 hours ago (around 6UTC), and OMC DCPD sum also seeing the same thing.
OMC_DCPD was saturating starting at 7:46UTC, and only ended when I'd completed the removal of whitening from the OMC using guardian.
Unsure of what to do now - calling to see what's wanted now:
Here's the violin mode medm: two modes on ETMY are now over 5
called and talked to Keita - turned violin mode damping on, raised some gains to help damping - attached:
snapshot of the range dropping and recovering - note that an EQ arrived just before I took the snapshot, and the range drop is seen on the far right
violins as of 15:00UTC (before Maintenance)
TITLE: 08/01 Owl Shift: 07:00-15:00 UTC (00:00-08:00 PST), all times posted in UTC
STATE of H1: Observing at 52Mpc
OUTGOING OPERATOR: Jim
CURRENT ENVIRONMENT:
Wind: 8mph Gusts, 6mph 5min avg
Primary useism: 0.01 μm/s
Secondary useism: 0.10 μm/s
QUICK SUMMARY: locked and in Observe
TITLE: 08/01 Eve Shift: 23:00-07:00 UTC (16:00-00:00 PST), all times posted in UTC
STATE of H1: Observing at 53Mpc
INCOMING OPERATOR: Cheryl
SHIFT SUMMARY: Mostly quiet
LOG:
0:10 NLN, with violin mode damping left off, per Sheila's log
0:50 IFO went to commissioning for no reason I could find. IFO guardian only showed some spm diffs on SEI ETMX(that had been there all along), but there were no SDF diffs. No one had touched anything.
Guardian DIAG_SDF suggests a difference showed up momentarily on SUSITMX
From DIAG_SDF log file:
2017-08-01T00:50:00.82615 DIAG_SDF [RUN_TESTS.run] USERMSG 0: DIFFS: susitmx: 1
Dataviewer second trend show the difference was only in effect for 3 seconds (00:50:02 - 00:50:04). Conlog did not report any settings changes at this time.
susitmx just threw us out of observe again at 0:08 UTC . DIAG_SDF logs one sdf diff, but I didn't catch it. susitmx guardian log shows nothing.
I'm running a python script on zotws6 which will print the name of the first channel in the difference list if the number of SUSITMX SDF diffs becomes non-zero.
Switched "carrier off" on the 2 ifr units at around four in the afternoon.
J. Kissel
Operators have been stifled with junk (bad/absent calibrations / wrong sensors / poor plotting) templates for violin modes that have been linked to the violin mode MEDM screen:
/opt/rtcds/userapps/release/isc/h1/scripts/
V_mode_hunting_NK_2nd.xml
V_mode_hunting_NK_3rd.xml
V_mode_hunting_NK.xml
for quite some time. I took the time this evening to
- Changed the start frequency of the template to be 0 Hz (where they had been set to 490 / 900 / 1400, i.e. the lower band of interested for each template). Maybe this is an old commissioner's tale, but I recall that if you don't set the start frequency to zero, then your ASD results get distorted.
- Restored the calibration of DELTAL_EXTERNAL so that it (a) is correct, and (b) matches the front wall. This was done by loading in the transfer function calibration from
/ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O2/H1/Scripts/ControlRoomCalib/caldeltal_calib.txt
- Added the OMC DCPDs to all templates, added advice on what the results need to be in order to use DC READOUT.
- Where possible (i.e. only on in the 500 Hz fundamental template), I plotted ASC_AS_C raw segments to check those for saturation. Added similar advice about results in order to use ENGAGE_SRC_ASC.
The templates have also no been committed to the svn, where they hadn't been before.
We have left the violin mode damping off for one lock, with the intention of measuring violin mode Qs to compare to what was measured in 23383. We had the damping off starting at 00:05:00 UTC, and will leave it off for the rest of this lock.
ETMY MODE 7 rang up, and possibly ETMY MODE 2. If this data is unusable, we can try this again but leave all the damping on for ETMY.
Following an e-mail exchange between Jeff K and Brett Shapiro about alog 37847, Daniel, Thomas Vo and I tried to follow Brett suggestion to swing ITMY and see if we could reverse the hysterisis.
First we took the ISI to isolated damped, then set the top mass damping gains to 0. Daniel applied a 0.3 Hz excitation to M0 OPTICALIGN_P until the top mass was swinging with an amplitude of more than 200 urad according to the test mass osems (the sift we saw was about 600 urad). He slowly ramped this down over 5 minutes, we let the suspension swing for a few more minutes. We don't see any shift in the pitch of the optic before and after this test.
The second attachment shows the same channels during the EQ. The motion durring the EQ was much larger than what we applied with this test, so it might be that we need to try swinging harder.
Here is the suggestion from Brett's e-mail:
If you imagine that the hysteresis is coming from dislocations in the wires all moving to one side of the wire or the other (front or back), then letting the quad swing and ring down slowly causes the dislocations to spread out evenly, washing the hysteresis away. But if the quad tips a lot, but rings down very fast, like from an earthquake causing it to bang hard into the stops, then the dislocations may pile up on one side, causing a static pitch. So the hysteresis plot I attached is really a worst case.
A simple thing to do to check if this hysteresis thing is the culprit is to get the pendulum swinging in pitch, with an amplitude at least as large as the offset you see (if possible). Then let it ring down without damping. If the offset goes away or gets smaller, then this was it. If not, then it must be something else.
A minor correction on what I said, which Sheila already caught, is that you would want the swing to be at least as big as the motion caused by the earthquake, not the remaining offset. If this causes the pendulum to bang into the stops, than it may not be feasible to do a proper swing.
Sheila, with many conversations with Keita, Jeff K, and others
We have several reasons to believe that something changed in our suspensions durring the Montana EQ. (See alog from Beverly Berger and Josh Smith, (37775) which got us started looking at this, and Cheyl's log about the large triples, (37674). We are still looking at some of the data from alignment sensors, but here arer some things we can say:
There is more to be done checked on here, for example, checking if this has happened at any other times during the run (I checked one large EQ, I see no shifts like this), checking yaw (which has much smaller shifts than pitch), checking the triples, and looking at the alignments of the reaction chains. Can we interpret the information we have to make a gues aout what might have changed in the suspension? Wire slipping or some kind of damage to the prisms are some things we have been thinking about.
Hysteresis is a possibility here. We discovered on the LASTI quad during one of the early builds that these suspensions have significant hysteresis in pitch. That is, if you tip the stages a given amount, they will not come back all the way, leaving you with a pitch offset. The attached plot shows a measurement of this effect from LASTI, showing a pretty standard looking hysteresis plot. We learned that you can 'undo' any offsets by getting the pendulum swinging, and letting it slowly damp itself. The slower the ringdown, the closer it returns to its nominal 'equilibrium' position.
The offsets you see here don't look any bigger than what we saw there, though granted we were trying to measure the effect, so pushed it pretty far. Then again, we didn't have any major earthquakes either.
There were many documents written to investigate what we saw at LASTI. Mark Barton's document, T0900103, includes a list of most if not all of them.
So it could be that this earthquake induced some hysteresis offset, or perhaps there was an offset already and the swinging motion from the earthquake removed it. Anyway, try swinging the pendulum in pitch with some large *but safe* amplitude, and you should return to the nominal 'equilibrium' position, if it isn't already there.
Looking at Beverly log (37775) that shows DC changes in the pitch offset across the earthquake time. Are the changes in the pitch in the lower masses compatible with the reported change at the top mass?
Here are some additional plots, for those who are interested in what happens to the osems between the reaction mass and the top mass. I also have plots that show torque applied to the reaction mass vs measured pitch, these aren't very useful because we don't change the torque applied to the reaction mass, but they do show that there were similar shifts in the reaction chain. In order to interpret the data from the L1 and L2 osems we will need to account for shifts in the reaction chain.
Posting a jpeg version of the LASTI hysteresis plot above, since the pdf was causing issues.
Also, here is a summary of the procedure I used to make the plot way back in 2008:
"These data points are separate pushes and releases. The procedure was to put the top mass on its stops with the rest of the chain suspended. Note, the quad was on the test stand outside the chamber at the time. Then the top mass stops were used to tip the top mass some amount in pitch. The angle of the test mass, with the top mass still tipped, was measured with either an autocollimator or optical lever. That test mass angle is the X axis in the plot, called ‘Input Pitch’. Then the top mass was released slowly to avoid oscillation, and the test mass pitch angle was recorded again. That value is the Y axis, called ‘Output pitch’. This process was repeated for successively larger and larger Input pitch values, until I was afraid to tip the suspension any more. I then started to tip the suspension in the other direction until I was again afraid to tip the suspension any more. And finally, to close the hysteresis loop, I repeated some of the data points along the original tipping direction."
