patrick.thomas@LIGO.ORG - posted 16:00, Thursday 23 January 2014 (9508)
Yuta, Koji puttiing equipment for WP 4409 back in squeezer bay
The h1dmt0 and h1dmt1 computers are in the process of being upgraded to Ubuntu 12.04. This will take a couple of hours.
Hugh, Sebastien
Position loops (UUG 5Hz) have been installed on HEPI-ITMX.
The design of those loops can be find in the SVN at:
/ligo/svncommon/SeiSVN/seismic/HEPI/H1/ITMX/Scripts/Control_Scripts/Version_5/GET_H1_ITMX_Controller.m
The function GET_H1_ITMX_Controller stores the zpk filters into a structure called FILTER_OUTPUT. This structure is used by the HEPI scripts steps 6&7 to generate the plots and saved the filters into the good format.
Sebastian and I looked at trends back during the day to set the Target positions for HEPI. We left the Pringle modes out of the loop for a while because they were generating large drives to the actuators. After a couple hours we saw the trends of the pringles not going anywhere fast so we set those targets and engaged those loops.
When I looked at longer term trends to make this log I see that the HEPI Location values were zero until Hugo and I worked on the position loops on the 16th and the input matrices were populated. The trends look very flat/stable until ~8pm 21 Jan. The ISC crew reports suddenly experiencing a sudden misalignment. They traced it to ITMx HEPI. Some confusion about whether or not they attempted to engage the uncommissioned HEPI loops but the point is, HEPI moved and stayed. They realigned the SUS and continued. So HEPI will stay here.
Attached is the current Cart_Bias for the Target positions. If you see the HP or VP numbers far from the Targets, know that these are AC coupled.
Done.
From ~ 2:00 - 2:10 PM
Done
Mark B.
The previous best HTTS model (20130723TipTilt) had two issues: (i) the MOI values from P1100090-v5 were obviously wrong and had had to be arbitrarily increased by about an order of magnitude to match the observed frequencies, and (ii) there was no allowance for built-in ECD. Bram supplied some new MOI/mass values which were more up-to-date and didn't have the divide-by-10 typo. And I added ECD support to the underlying Matlab.
The result is a new parameter set, 20140123HTTSdamp (Mathematica) or httsopt_damp.m r6162 (Matlab). See
https://awiki.ligo-wa.caltech.edu/aLIGO/Suspensions/OpsManual/HTTS/Models/20140123HTTSdamp and ^/trunk/Common/MatlabTools/SingleModel_Production/httsopt_damp.m.
The mass and MOIs are now as-built and the damping parameters have been chosen to roughly reproduce the observed peakiness in TF data from H1:RM1 on 12/20/13 (see attached plots for L, P and Y). The match to the mode frequency is now very good without any fudging.
Done
The Green Team has been been complaining of excess ~0.5 [Hz] pitch motion of the QUADs for the past few days. We had first chalked this up to the fact that the ISI were not yet at full performance. However, after Sebastien and company got both chambers up to a reasonable performance, and even going so far as to tailor the blend filters to get as much performance at 0.4-0.5 [Hz], the cavity pitch motion was deemed still too high. I finally just took a comparison ASD of both chambers and their optics' optical levers, and discovered a giant spike in ISI ITMX at 0.5 [Hz], and a subsequent comb cascading up to higher frequency. The 0.5 [Hz] (and its harmonics -- note that 0.5 [Hz] is *not* the fundamental) spike appears obviously in both the ISI ITMX ST2 GS-13s. Only the 0.5 [Hz] harmomic shows up in the SUS ITMX optical levers -- but it happens to perfectly coincide with the QUAD's first "pitch" mode at 0.5 [Hz]. Awesome. Given that we've seen similar combs at LLO (and I believe DetChar has confirmed their presence at LHO), which were confirmed to be a result of the BSC-ISI's capacitive position sensor master oscillators beating against each other, I immediately suspect that this is our problem. In order to confirm, we're going to run the corner BSCs in various configurations of OFF to assess whether the comb disappears, if not at least changes. Note: since this is systematic problem seen at both sites, a permanent fix to the problem is in the works (i.e. syncing all chambers to a single, GPS synced, master oscillator). It's just not ready yet.
With the recent cabling work at EY, wanted to record notes of where we're at and inventory.
Problem Cable
Performed elec short tests on in-vac cables, and found an issue with one: D1000223 S1104078's pin #1 & #23 (may be able to use a D1000225 in its place)
Screws On Cable Feedthru Connectors Confirmed To Be "Screwed In" All The Way
It has been brought up that a possible source of grounding are the sets of screws on the connector which attaches to a Chamber feedthru. Through handling sometimes these screws can unscrew, and even a little bit will have them protrude out and short to the Chamber. I went ahead and applied a torque to all of these connectors for TMS cables. (attaching a photo of these screws on our 25-pin connetors....with head on shot and profile)
Missing Screws For Attaching To Feedthru
In the table below there is a note of cables which are missing screws which attach cables to the in-vac side of the feed-thru.
Cables Screwed Together & Unfortunate Cable Bracket Attachement
Since we've tested all these cables, I went ahead and screwed all cables together (noting we'll have to unscrew the bad D1000223 at some point). In our Cable Layout drawing (D1200111), the D1000921 cables for the Picomotor & Beam Diverter are said to be OK with connectors "with or w/o metal ears". If you do not have metal ears, you sort of make life hard for you in that if you have to look at a cable you can't simply unscrew it from the cable bracket. You remove access to the Cable set screw, so you have to remove both cables and then disconnect cables---which is a pain.
ICS & Inventory Of Cables
I have made an Assembly Load for the TMS EY Cables (ASSY-D1200111).
|
In-Air Cable |
Chamber feed-thru |
Seismically-Responsible Cables | Cable Bracket | In-Vac Cable |
Cable Bracket on TMS |
In-Vac Component |
|---|---|---|---|---|---|---|
| : Not sure of name : | E6-7C1 |
D1000225 s/n S1104782 *no feedthru screws? |
CB-5 , 1st Floor | D1000234 s/n96-911 | no CB | OSEMS: Face1, Face2, Face3, Left |
| : Not sure of name : | E6-7C2 |
D1000225 s/n S1104778 *no feedthru screws? |
CB-5 , 2nd Floor | D1000234 s/n96-901 | no CB | OSEMS: Right, Side, ---, --- |
| : Not sure of name : | F2-2C1 |
D1000924 s/n S1104106 *no feedthru screws? |
CB2 , 1st Floor (was 3rd Floor of BSC6 CB) | D1000568 s/nS1104468 | CB-primary, 1st floor | Green QPD (D1000231 s/n C2Q001 aka S1400085) |
| : Not sure of name : | F2-1C2 | D1000924 s/n S1104469 | CB2 , 2nd Floor (was 2nd Floor of BSC6 CB) | D1000568 s/nS1104109 | CB-primary, 2nd floor | Red QPD (D1000231 s/n C4Q001 aka D1400087) |
| : Not sure of name : | F2-1C1 |
D1000223 s/n S1104077 *no feedthru screws? |
CB1 , 1st Floor (was 1st Floor of BSC6 CB) | D1000921 s/nS1104116 | CB-entry, 2nd floor | Picomotors (D1000238 s/n1104586) |
| : Not sure of name : | F2-2C2 |
D1000223 s/n S1104078--possible short w/ pin 1 & 23 |
CB1 , 2nd Floor (was 4th Floor of BSC6 CB) | D1000921 s/nS1104114 | CB-entry, 1st floor | Beam Diverter (D1000237 s/nS1104289) |
Daniel, Alexa, Sheila
We have tried to measure the green modulation depth of the X arm by moving the modulation frequency (24.410610MHz) so that the sidebands are antiresonant in the cavity, and then locking the cavity on the carrier (+polarity on servo board) then the sidebands. We used H1:ALS-C_COMM_A_LF_OUTPUT to measure the transmitted power.
| carrier | sideband | |
| no attenuator | 491.3 | 184 |
| 6dB | 620.34 | 81.67 |
| 12dB | 664 | hard to measure |
I have changed the offset to -66 counts, so that with no light on the PD we get zero counts (-0.26 counts) (someone had changed this from the -61 I originally set to -100 sometime last week, I am not sure why). With the single shot beam we get 46.06 counts.
Below are plots ot the transmitted powers (normalized) when locked on the carrier and sideband, and the second plot is their ratio. I used the ratio to do a simple fit for the modulation depth that we started with, and got 0.80 Now the modulation frequency is restored to the nominal 24.407079MHz and the 12dB attenuator is in place so our modulation depth should be 0.2.
Also, the ratio of the transmitted power on resonance/ single shot beam should be:
1/(1-sqrt(R1*R2))^2=22.6 for a perfectly mode matched carrier with R1=0.63, R2=0.99, taking into account the 0.2 modulation depth we would expect 22.17. Since we curently measure a ratio of 14.4 between the single shot and locked trans power, which is 65% of what we would expect. This means we have about 45% of our power in higher order modes.
Daneil's alog about the RF powers is 9484
Done with install, Gerardo opening GV7
GV7 open
For WP 4409
Work done, VEA transitioned back to laser SAFE.
The h1nds0 daqd will be shut down for up to 10 minutes to build RCG 2.8.2 versions of daqd to support longer frames. Longer frame collection should be implemented about 11:00 PST for h1fw0. Following about an hour of testing, and assuming testing is successful, longer frame writing will be implemented on h1fw1 about 12:00 PST.
Due to frame writer 0 crashing 35 minutes after starting while writing longer frames, we are delaying the change to framewriter 1 until 13:00 PST at the earliest.
[Jeff K, Stefan B, + detchar] The bitmask defining the summary bit for the Input Mode Cleaner state-vector (ODC) has been trimmed to respond to only those bits recording the ASC WFS switch (bit 1), and the relative power indicators (bits 11 and 12). In this mode glitches in the WFS alignment will be recorded by the relevant bits 2-9 but won't affect the summary bit, and a low-power mode will still result in summary=GOOD. This was tested first at LLO and logged here. This hasn't been set in the safe.snap, but will be monitored and coordinated over the next few days.
We confirmed that when the WFS alignment (mainly DOF2 Pitch and DOF1 Yaw) strays outside of the thresholds set, it does not have a noticeable affect on the calibrated length of the cavity. Attached are ASDs for both LHO and LLO of IMC_X_DQ. Plotted are two nominal times (red/blue lines) when everything is aligned (all the ODC bits are green) and two times (yellow/green) when the WFS are reporting poor alignment (the y-axis is not in counts/rtHz but the units of the channel - m/rtHz (?)). There is no noticeable difference between the four traces on a given plot. For the time being, we feel ok in removing the WFS alignments from the summary bit (as reported) however we plan to do long term studies of what the thresholds for each DOF should be and maybe then they can be re-added to the summary bit. Question for whoever can answer this - why is LHO so featureless (compared to LLO) and why are the scales (y-axis) so different between sites? Thanks
