sheila.dwyer@LIGO.ORG - posted 14:13, Wednesday 16 April 2014 (11390)
PSL back to science mode
Scott & I got the last Horizontal L4C leveled and then a few more Actuators attached. Should complete this initial install tomorrow morning. IAS in the afternoon. I say initial because commissioning may find some details that need addressing.
Since the Faraday swap on monday, we have more green power.
On the table we measured 50mW going towards the chamber, where previously we had measured 47mW (6% increase). On QPD A we had a 10% increase in the sum, on QPD B a 30% increase in the sum.
Alexa and I measured 19uW arriving on ISCT1, compared to 12 uW measured in alog 9191 ( 60% increase from January)
We also noticed that the beam quality was better than it had been. We measured some profiles, it is a guassian beam, they will be attached to this post later.
When the arm is locked we see a 35% increase in the transmitted power measured by the COMM BBPD, compared to last week.
We measured the phase noise of the low noise VCO. We used the COMM PLL which deploys a single stage frequency-difference divider (FDD). The PLL was locked to a fixed frequency oscillator (ifr locked to GPS running at 78.95 MHz). The output filters of the PLL were: boost engaged, generic disabled, VCO comp off and low pass off. We added a SR560 with 3 kHz low pass and a gain of 10 to clean up the high frequency part. We then looked at the error point of the common mode board (H1:LSC-REFL_SERVO_ERROR). The calibration is as follows:
25kHz/V (VCO sensitivity) / 25 (boost gain) * 5 (gain divider in VCO output path) / 10 (SR560 gain) / 100 (gain in ERROR readback) * 1V/3200cts = 0.0016 V/cts
For the SSB (single-sided sidebands) we divided by an other sqrt(2). This plot matches nicely to the high frequency SSB curve measured in T0900451. Above 100 Hz we see some excess phase noise which is most likely due to the ifr. The two curves together characterize the low noise VCO all the way from 1 mHz up to 100 kHz. To get the phase noise of the VCO without the FDD we can just multiply the attached curve by 10.
The rms frequency noise of the VCO was previously measured in alogs 6972 and 6865. The numbers were 2 Hz rms for a VCO with FDD and 16 Hz for a VCO without. This is in good agreement with the current measurement.
Although it seems the watchdog trips at the ITMs are understood, there is now another variable to consider. The cleanroom that has been inserted into the beer garden for the upcoming ITM swaps has some of it's curtains draped over the HEPI crossbeams. With the cleanroom not currently running and no plans to move the cleanroom in the near future, the interference should be rather static. See pics for a reference to the extent of the interference (and let us know if it is a problem).
Possibly due to craning in the LVEA.
Tripped April 15 17:10:34 PDT.
model restarts logged for Tue 15/Apr/2014
2014_04_15 11:35 h1hpiham4
2014_04_15 11:38 h1pemcs
2014_04_15 12:15 h1broadcast0
2014_04_15 12:15 h1dc0
2014_04_15 12:15 h1fw1
2014_04_15 12:15 h1nds0
2014_04_15 12:15 h1nds1
2014_04_15 12:16 h1broadcast0
2014_04_15 12:16 h1fw0
2014_04_15 12:20 h1hpiham5
2014_04_15 12:23 h1hpiham2
2014_04_15 12:23 h1hpiham3
2014_04_15 12:26 h1hpibs
2014_04_15 12:26 h1hpiitmx
2014_04_15 12:26 h1hpiitmy
2014_04_15 12:28 h1hpietmx
2014_04_15 12:28 h1hpietmy
2014_04_15 12:29 h1hpiham6
2014_04_15 16:39 h1susquadtst
2014_04_15 21:03 h1fw1
A busy maintenance day. All restarts expected except the second start of h1broadcast0 and the later restart of h1fw1.
J. Kissel, A. Pele
We've begun the clean up and unification of design for optical lever damping that has begun to slowly and differently percolated through suspension types and sites. Arnaud has created a library part called
${userapps}/release/sus/common/models/FOUROSEM_STAGE_MASTER_OPLEV.mdl
a week or so ago which has been used in the modified
${userapps}/release/sus/common/models/QUAD_MASTER.mdl
part which covers a generic QUAD suspension. However, there were still a few bugs in that library part which I fixed up:
- breaking a vestigial link to the beam splitter master library part,
- connecting the ODC damping status up and out to the top master level adding two new bits
We've not compiled anything and the MASTER model parts remain uncommitted. We'll compile, install, restart, and restore, and then clean up MEDM screens tomorrow.
Note that we are cleaning up this stuff so that we can make progress on all of the ECRs that HAVE been approved, including watchdog clean up, adding EPIC storage of alignment offsets, adding dither alignment paths, and adding BSFM M1 to M2 stage damping.
Debate over whether to include optical lever damping in DRIVEALIGN path (which contains frequency dependent alignment decoupling [say if the P2Y coupling is large] and the plant compensation filters [to convert complicated P2P and Y2Y transfer functions into single-pendulum-like transfer functions]):
- Currently QUADs feed optical levers through DRIVEALIGN matrix, motived with the argument that "we shouldn't have 10 different paths to actuate on a stage. If we put the effort into 'diagonalization' of one path, it should apply to all other control loops."
- The BSFM does NOT feed through, because it was put in quickly before any such discussion was had, or any effort was put in to 'diagonalization'
- The HLTSs currently have no optical lever damping, but LLO commissioners have requested it.
- If we use Arnaud's FOUROSEM_STAGE_MASTER_OPLEV for all three (QUAD, BSFM, and HLTS) suspensions' lower, FOUR OSEM stages, then we'd only have one library part to maintain. Good.
- HOWEVER, we've already approved to have M1 to M2 stage damping on the BSFM, which means it's M2 (FOUROSEM) stage has to stay unique.
Gosh darn it. Note LHO hasn't measured if we see the same SNR and damping improvement that LLO got from this unothordox extra damping...
DavidH, Greg, Thomas TCSX: Plumbing sorted (except for one last connection to water-cooled beam dump, awaiting fittings), optics mounted and optical mounts are positioned on the table ready for alignment. Optics that are not within table are accounted for. Last few cables are to be connected. TCSY: Enclosure re-cleaned, plumbing connected up (except for beam dump) and water manifolds in final positions, laser/RF driver cables are all dressed and connected to feed-throughs. See images for current TCSX/Y table layouts
Have we located the keys for the table door locks?
They were found in the enclosure after David was able to tidy up.
Late start but this morning I lifted the HEPI with it's DSCW Springs to pull the elevation within spec: -325.7mm vs -325.3 so less than 0.5mm low. The Dial Indicators show that the position remains within spec.
After that, I got the first Horz L4C Leveled and then the first vertical Actuator attached. After lunch (1400) Scott joined me and we leveled another L4C and attached the second vertical Actuator. The 3rd L4C is leveled now and we'll continue tomorrow.
Departed the LVEA ~1645pdt.
ITMX and ETMX ISIs are damping now, so that seismic people (Rich) can get an input spectrum on a windy day. I will be walking around the end station optics table, but quietly. Sheila
Don't know why it tripped
[Sheila, Kiwamu]
Sheila had a difficulty in relocking the IMC. So I looked into it to figure out what is happening and fixed it by steering the PZT pit.
It was apparently due to a misalginment, mainly in pitch according to the GigE views. Looking at one-day trend, I found that the PZT pit suddenly had changed its value from 400 to -240 counts at around 8:20 am local. A mysterious point is that , after the PZT jumped, the IMC was still able to lock itself for approximately 4 hours.Then it ran into a situation where it was not able to hit the LSC trigger. I didn't see a significant drift in any of the MC suspensions. I fixed it by steering the pitch of PZT back to +300 counts. It relocked right away and we ran the offload script so that it doesn't loose the good alignment by a unclok. We should keep our eyes on this PZT.
The attached is the one day trend.
I changed the PZT offsets this morning to account for the temperature drop in the laser room over night. Robert then increased the nominal, so it is expected that the alignment shifts back towards the older settings. The correct way is to use the centroid of the reflected beam when not locked and steer the PZT to the same position. Dave added theses channels to the frame, so they can be trended (if you can figure out how they are named). We probably should add an error condition to the IMC locker which checks this at the beginning.
Here are spectra of the ETMX and ITMX oplevs. All blends are on Tbetter, sensor correction was on for ETMX and off for ITMX.
We are bothered right now by large amplitude motion (around 1urad pp) in yaw at very low frequencies.
We've seen this on ITMY before and changing out the laser seemed to fix the problem. I've attached some graphs indicating more clearly what might be occurring. The time series indicates that the power is dropping on all quadrants in with 20 second intervals; although this is very difficult to see by looking at the normalized pit and yaw signals,you can see it very clearly when looking at the individual segments. Similarly, in the spectra, there is only a small indication at 1.0-2.0 Hz that something weird may be going on but when we look at an individual segment, you can see weird spikes starting at 0.056 Hz and then the harmonics following through to higher frequencies. For comparison, I included the ETMX Segment1 spectra and you can see that we shouldn't see those spikes. I'll be investigating this today but my first guess would be either the laser or laser power that's causing a dip in power every 20 seconds, which is pretty odd because we got this laser new "off the shelf" from microlasers.
Some improvement but still oscillatiing a bit.
I've attached new spectra comparing the WIT sensors and the optical lever to show that we can see real motion of the optic for the coil balancing procedure that the SUS team is trying to perform. Also attached is the comparison between ETMX and ETMY, although the ISIs are in a different state, we can see that they are comparable. The last image attached is a trend showing that we still see this power oscillation with a peak to peak of about 240 counts with a period of ~3 seconds. This is much better than the p-to-p of 2500 counts every 19-20 seconds in the previous laser configuration.
Trying to hunt down the exact noise source proved to take longer than I thought it would. I ended up switching between three variables and chooising the best configuration: The laser, the power supply, and the outlet. We have not switched this laser to run with the power board yet but once we implment this change, it'd be interesting to see if this oscillation goes away or gets worse. Richard thinks that the noise comes from the thermo-electric coolers in the laser itself, in which case switching to the power board probably will probably not change this artifact.
The hydrogen outgassing rate measured is 75% of that measured in 2000. The atmospheric accumulation is about 6.9 x 10^-8 torr liters/sec and seems high relative to that in y1 and the accumulations made at LLO. The value suggests a small leak. If this is in the beamtube it is at the threshold of not being able to be found with our current techniques. The results of the accumulation were more difficult to calculate and have more uncertainty due to the method of connecting the RGA to the beamtube. The connection was made by a corrugated small diameter tube rather than a mount directly on the beamtube. The pumping speed of the tube and the pumping speed of the RGA as an ion pump need to be accounted for. The attached pdf file shows the influence of the connecting tube and presents the results.
Yep, this level of air leak would be challenging to locate by current techniques. It's probably small enough not to directly impact sensitivity, but it seems important to establish whether there's a degradation process at work, like LLO Y. We should consider a repeat accumulation, perhaps with three sampling points along the 2km module.
I estimate some 3000 cm^2 of viton in the large valves exposed to this volume. This may account for some of the air. Anyone know the outgassing rate of viton after 140,000 hours under vacuum?
We should probably reconfigure the RGA to eliminate local orings and the low conductance of the flex hose. Kyle and I felt we had nothing to loose by trying the quick and dirty method first. Practice makes better.
Correction on the surface of viton contained in the LN pumps and large gate valves.
There are 4 gate orings entirely in the vacuum (45 inch diam) and 2 gate orings 1/2 exposed at the closed gate valves(45inch).
There are 4 flange orings 1/2 exposed where the gate valves bolt to the ln pumps(45 inch diam)
There are 2 bonnet seal orings 1/2 exposed in the two gate valves open to the volume - each is 144 inch long.
Total surface area ~ 6970 cm^2
If the entire air leak is allocated to this we get an outgassing rate for the viton of 1e^-11 tl/sec/cm^2.
Last friday, Betsy and Travis added weight to the middle mass of SRM currently on the test stand, in order to center the flags with the osems cf alog 8367. A first round of TF have been taken before that, showing clean measurements. A new set have been taken monday night to assess that there's no major change. The new results are showing an excellent match with the previous ones
Results in attachments are described below :
(1) H1 SUS SRM M1-M1 (top mass) comparison between modeled and measured undamped transfer functions for the 6 dofs
(2) H1 SUS SRM M1-M1 comparison between modeled and measured damped transfer function for the 6 dofs
(3) H1 SUS SRM M1-M1 undamped TF comparison before and after adding weight to the middle mass
data have been commited under the svn
I meant SRM in the title
For acceptance review I also added the spectra of SRM during phase 2b (currently plugged on the test bench).
The first 3 pdfs are showing the effect of the top mass damping on each stage. The last two pdfs are comparing SRM spectra with other hsts suspensions (PRM and SR2) during phase 2b, damping off and damping on.
Everything looks good for all osem signals, damping is working fine by significantly reducing the Qs of the resonances, and both damping on and off spectra are similar to other HSTSs at the same 2b stage of testing.
MC1 MC2 and MC3 Phase 3b M1 to M1 transfer functions have been measured over the past weekend. They all show good agreement with model and previous measurements.
The attached files are showing comparison between model, phase 3a and 3b (except for MC2 3b vs 3b)
MC1 (allhstss_2013-05-29_Phase3b_H1MC1_ALL_TFs.pdf)
MC2 (allhsts_2013-05-29_Phase3b_H1MC2_ALL_TFs.pdf)
MC3 (allhsts_2013-05-29_Phase3b_H1MC3_ALL_TFs.pdf)
files and data have been commited on the svn under the following directories :
/ligo/svncommon/SusSVN/sus/trunk/HSTS/Common/{MatlabTools/Data}
/ligo/svncommon/SusSVN/sus/trunk/HSTS/H1/${MC1/MC2/MC3}/SAGM1/${Data/Results}
All of those top stage transfer functions look good.
The only thing to note is that MC2 has its pitch (1Hz) and roll modes (1.5Hz) coupled into the vertical mode as seen on the 3rd page of the 2nd pdf. It is significantly reduced by the damping
This coupling increased after pump down (orange/pink vs cyan/green).
