edmond.merilh@LIGO.ORG - posted 09:15, Monday 14 May 2018 (41966)
PSL Weekly Report - 10 Day Trends FAMIS #6200
Images attached to this report
ISC
- Beam to HAM6, AS AIR check, AS_C, and lock DRMI
- SRM AWC testing on going. SR3 heater measurements happened over the weekend
PSL
- "Wrap up" this week
SUS
- Zero ITM OpLevs and run ITM TFs
VAC
- MY work continues
- Ion pump installs not till next week
EX
- TMS and ETMX work continues
- AMD install end of the week
- PCAL baffle install this week
Other
- HAM6 Door will be coming off this week
- Access control work will be in LSB today
Tuesday Maintenance
- RF balun investigation
FAMIS7490
Laser Status:
SysStat is good
Front End Power is -0.003303W (should be around 30 W)
HPO Output Power is 0.7854W
Front End Watch is GREEN
HPO Watch is RED
PMC:
It has been locked 3 days, 15 hr 28 minutes (should be days/weeks)
Reflected power = 16.46Watts
Transmitted power = 51.93Watts
PowerSum = 68.39Watts.
FSS:
It has been locked for 3 days 15 hr and 28 min (should be days/weeks)
TPD[V] = 1.073V (min 0.9V)
ISS:
The diffracted power is around 1.8% (should be 3-5%)
Last saturation event was 3 days 15 hours and 28 minutes ago (should be days/weeks)
Possible Issues:
Front End Power is Low
ISS diffracted power is Low
LRA out of range, see SYSSTAT.adl
This script definitely needs to be updated. I will talk to the PSL crew.
Thomas, Dan Brown
This morning we wheeled out the SRM heater to the north side of HAM5. The CO2 laser cannot be switched on currently until the SOP is complete, the controller for the CO2 is on Thomas' desk. The goal today was to try and align the table to the two steering mirrors and the holes in the SRM baffle. We managed to center on the ZnSe viewport and the first steering mirror. Changing pitch and yaw there we could trace the beam around the edge of the first baffle hole, we were unable to do the same on the output baffle hole. When we believed we were centered correctly the alignment beam could be clearly seen on the baffle on the HR side of the SRM, we're currently unsure whether that should be the case or not. We also do not have a good view of the beam dump from the north side which makes trying to align into it difficult. We still have some alignment work to do here before we are ready.
The north HAM5 door area is now sectioned off with screens between the squeezer table and the clean room, pictures of the area are attached.
Eddie, Corey
Eddie exported the AWC mirrors and beam dump from the SolidWorks model so that I could add them to the Zemax model. The attached .pdf shows the path of the CO2 beam into HAM5.
Aidan confirmed that most of the power will be absorbed by the optic, so only the beam labeld 'AR Reflection' will be present with the actual CO2 beam. I included the other beams (transmitted, HR reflection) in the model since the alignment laser is not absorbed, and thought it may aid with alignment.
Betsy and Jason conducted the initial alignment as documented in:
LHO aLOG 39669 https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=39669
Is there any reason that this successful initial alignment would not be easily recaptured? Seems that there should be enough adjustment possible to sort this out.
Not really any major issues. It's just awkward to check the alignment using the reflections on the SRM baffles: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=41724. We asked Corey to check where we should see spots on the SRM HR baffle so we had a better idea of what we were looking at.
We managed to get the CO2 beam on the SRM and induce a lens using this method yesterday: https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=42254
Alexei, Dan Brown
We did some beam profiling measurements of the SRM heater to determine how far away the table should be positioned from the ZnSe viewport and if the beam sizes are correct. To take beam profiles of the CO2 we used a knife edge and a power meter. The measurements were compared to a simple model of the setup which matches up well. The aim was to get around a 5mm radius at the SRM, which we can achieve if we position the table so the top of the periscope is about 1m from viewport.
Measurements and analysis can be found in:
https://git.ligo.org/IFOsim/Finesse_playground/blob/master/daniel_brown/LHO/CO2_knife_edge.ipynb
Daniel, Jenne, Fil, TVo
In preparation for this weekend's work to try to do an OMC Scan measurement while varying SR3 RoC, I decided to try to make sure our new alignment of the IMC could make it down to HAM6. I was able to get light on AS_A_DC and AS_B_DC and close the centering loops but I noticed that AS_C had no light on it which is very odd.
All the electronics were turned back on from our grounding loop searching from April 26-27 however, the attached trend shows that something dramatic happened to that QPD segments around Apr 27 2018 15:53:41 PDT which caused the individual segments to get much noisier and the sum to drop dramatically. I'm pretty sure we were laser safe at the time so it wasn't like the a beam falling off the diode. So I thought this might be some sort of electronics problem like a disconnected cable in the midst of the madness, but everything in the rack to the chamber feed-through seemed to be connected properly, Fil and Daniel tested the in vacuum QPD chain from the outside with a multi-meter and confirmed that there is a connected diode in chamber.
So maybe we're just not seeing light? Again, this seems strange because we see light on both AS_A and AS_B.
In general, this won't deter our work this weekend but we might want to get this figured out before we open the doors early next week or put it on the task list for the chamber. One method we haven't tried is looking for AS_AIR and seeing if that comes out of the viewport. Also, there is a picomotor we can actuate to steer the beam onto AS_C, but I'm holding off on this unless we know what is going on.
Made a quick check that the downstream electronics works:
Thomas, Dan
We checked the AS air and OMC transmission beams coming out of the north HAM6 viewports. They were too dim to see with cards but visible with the IR viewer, both are clearing the viewports fine
After more looking in the viewport with Keita, Daniel and TVo (including turning the illuminator on/off, and sadly seeing no response on AS_C, although we did see a tiny response on the OMC QPDs which are inside the shroud), I worked on picomotoring the AS_C steering mirror, in a last-ditch effort to see if we could ever see a signal. The answer so far is no. Using the largest picomotor steps, I went +/- 6 steps in pitch and +/- 10 steps in yaw from the previous nominal. I thought I saw a beam briefly, but it turned out to be something upstream, since it affected all of the AS diodes (including the WFS, which aren't affected by this steering mirror). I did a rough raster scan of the picomotor actuation, mapping out a rectangle. When I was finished, I put the picomotor back to the place that I started, so even though it won't have gone precisely to the old pointing, it should be close.
AS_C seems to have a lot of 60 Hz + harmonics noise, more so than the OMC QPDs do, when the IMC is offline (MC2 misaligned). Daniel points out that this shouldn't affect our ability to see a DC response, and the rest of the electronics chain passed some basic checks, but it does seem like a concern.
Daniel pointed out to me that the picomotor to HAM6 likely wasn't connected since ISCT6 isn't really back in it's final position yet. And, he was right. So, today I connected the HAM6 picomotor controller using some 10' DB25 extension cables, and then re-did my raster of the AS_C pico. Step size of "magnum", 10 x 10 grid. I didn't see anything at all that looked promising. The picomotor is back nominally where I started, although it's of course not precise. I think we're going to have to pull the north door of HAM6 tomorrow, as we discussed was a possibility at this morning's meeting.
In case you find a problem with the QPD diode, the head, or the invac cable:
The cable currently installed for AS_C is D1101654 / S1202409 (40"). It has only one head (Head #29). The diode installed is Q3000 InGaAs QPD #18.
cf. LHO ALOG 12244
There is only one spare InGaAs QPD diode (#15) left as an LHO spare.
The spare 40" cable is D1101654 / S1301546. This cable has no head. Therefore, to use this cable, the head needs to be transplanted from S1202409 or D1000231 S1202412 (36"), which has two heads (#34 and #35) and two diodes (#55 and #76) attached.
See also: E1101174
The two attached plots show before and after adding the final HEPI Y to RZ Cross Coupling factor.
This was first looked in alog 40106, 40117, & 40150 for ETMX. I want to look at ETMX again as that was before I looked at the geometric consideration I posted earlier in 41955.
I feel pretty good about the solution here though now.
The optical lever out put during the HEPI length stroke is the sum of actual optic yaw plus the expected apparent yaw due to the OpLev Length to Yaw Cross Coupling: OLOutput = Actual + Expected.
So I want to minimize the Actual = Output - Expected. Assume in this case the Actual is all caused by the HEPI Cross Coupling. See the 40106 alog for the conversions and such to get the units, and the amplitudes of the OpLev Cross Coupling factors are from E1200836. The signs of these corrections are subject to geometric analysis.
Anyway, with a CC factor for HEPI Y to RZ of 0.0065 in H1:HPI-ETMY_IPSALIGN_3_2 of the IPS Alignment Matrix, the residual is pretty much zero'd out as seen in the 2nd plot although I might be able to tweak it a little bit more.
This length drive on the ETMY HEPI is 0.8mm so a huge drive wrt anything we'd ever need to drive but why not get rid of it.
Re; "I want to look at ETMX again as that was before I looked at the geometric consideration I posted earlier in 41955." I've looked at that again in light of the geometry and conclude my result is valid for ETMX. The bottom line for this exercise is that the OpLev Length to Angle CC factor is Positive for ETMX (hence my correctness) and it is negative for ETMY. This is because the PD is on the opposite side of the beam tube and the beam is displaced in the opposite direction with a positive path length change. Thank you for your support.
Went ahead and did the ETMY X to RZ HEPI diagonalization. A value of -.0064 for H1:HPI-ETMY_IPSALIGN_3_1 does a thorough job of minimizing the cross-coupling. This has been accepted into SDF.
The butterfly valve for the purge air on the X beam manifold was missing its stop pin, allowing it to turn freely (more than a 1/4 of a turn), removed it and replace it with a spare on hand.
Upon further inspection of the bad valve we found the rolled pin was sheared off, also noted one of the two screws missing that fasten the disc to the stem, I was not able to locate the missing screw.
All of the IR alignment equipment was in use somewhere so I was unable to check alignment on the CO2 tables. Because of this the AOMs were left on the table so their absence wouldn't disrupt the beam path. Hopefully when the spares that are from the eLIGO days are repaired and upgraded, they can be used to replace the AOMs currently on the tables. The drivers have been removed and will be going to IntraAction with the AOMs. CO2X had switched channels on the flipper sensors, which would show the incorrect mask as being in place on the MEDM and guardian screen, and weird conditions to happen when trying to flip the masks. This could also lead to bad conditions like having both masks in the beam path.
There is also a guardian script in progress for requesting the different mask states and checking to see if the requested state is the actual state. There is still some work needed to make sure it faults to a reasonable state, and that it handles restarts in a way that makes sense to an operator. While working on this TJ and I noticed a weird glitch that can happen while flipping the central mask, the annulus sensor would switch states briefly even when not requested. I suspect that maybe the annulus sensor sees the central mask briefly while it is flipping. This currently doesn't have an impact on the control scripts, but could probably be fixed with more careful positioning of the sensors.
The beam dumps have been left on both CO2X and Y just before the periscope, so those should be removed unless a placebo test of TCS is desired.
Similar to what was done in alog 40394, we needed to shift the ALS laser in the Y end stations by ~16.2GHz (-5.39V on the crystal temperature control voltage) to bring it close to the PSL. This offset was then relieved using the front panel potentiometer. The crystal temperature now reads 29.4°C.
The ALS autolocker is working again.
E1200836 lists all the Optical Lever Length to Angle cross-couplings for the detector optics. All these factors are unsigned and so one might assume they are all positive. But the sign depends on what is moving and therefore what is the sign of the path length change. E1200836 indicates no path length change sign and as it geometrically dependent, it would be unreasonable for them to all be positive. One might assume the length change is from the Optic moving, not the detector or laser; that make sense to me. So to determine the sign is just a geometric exercise, right up my alley. A quick look at D0901467 confirms that the local coordinates in E1200836 used to position things are signed like the global system (what I assumed) and so if the Optic is displaced +Y and the path length change therefore is also positive, the beam on the detector moves -X. This displacement on the detector is equivalent to -RZ (-Yaw) and this is why I say the sign of the ETMY L2Y factor should be negative. A quick visualization for the pitch x-coupling would have one conclude the sign is positive but that depends because the HR normal is pointing -Y. The sign is important as the OpLev length to yaw coupling could be hiding the HEPI to RZ coupling.
To confirm that the sign for RZ is as I say, I drove the HEPI in RZ and looked at the Oplev. See T1000388 for the details of the SEI team (FabriceM) getting our local to cartesian transformations into the LIGO global coordinate system. The first graph attached shows a 100urad stroke on HEPI and the Oplevs output. The YAW response of the Oplev is correct in that as the HEPI yaws+ so does the optic according to OpLev. The PITCH response here is just the HEPI x-coupling to tilts so can't say anything about that from this plot. But, while a Y stroke may have some tilting (see the ISO_RX output,) the 3rd attachment shows the OpLev PITCH sign is incorrect when the Y is stroked: see in your mind's eye the Optic moving away from the low mounted OpLev LASER and the beam reaching the higher detector. Clearly the beam moves up, and I think of this as the Optic pitching up. However, since the normal to this HR surface is pointing toward the -Y direction, maybe to the ISC etc group, this is a negative pitch.
Regardless of the sign of pitch on the optical lever or the sign of the length to pitch cross coupling on E1200836, HEPI can't do anything about Optic Pitch directly so I'll don't need to know it.
Betsy and Travis told us that we can just use the top blade to rotate TMS is we're talking about O(100urad) instead of O(many mrad), so we did just that.
The procedure was: Remove ERM protection cover, adjust EX bias so the optic is centered in oplev, inject green laser, set the TMSX offset to get the retro-reflection on the table. Then loosen the blade (which makes TMS sag so no retro-reflection), turn the adjustment screw, tighten the blade, look where the beam goes, repeat.
It turns out that 1/4 turn of the adjustment screw took out 400-something urad, then Travis tried 1/16 of a turn and that took out the rest.
In the end TMSX alignment sliders were set to [P,Y]=[-74,5] urad, almost perfect for Y, and P is no worse than yesterday where it was [91,-585].
We also adjusted BOSEM depth. For F1, LF and RT we recently measured the open values and set the depth (41675) so didn't bother to remeasure open values but we reset the depth anyway as RT seemed to have drifted out.
For F2, F3 and SD, old and new offset as well as open value as of now are:
| Open today | New offset | Old offset | Reading before adjustment | Reading after adjustment | |
| LF | NA | NA | -9350 | ~9.9k | ~9k |
| RT | NA | NA | -13475 | ~16k | ~12k |
| F1 | NA | NA | -10515 | ~10.9k | ~10.7k |
| F2 | 18.1k | -9050 | -11164 | ~11k | ~9k |
| F3 | 23.2k | -11600 | -14150 | ~18.5k | ~11.3k |
| SD | 22.1k | -11050 | -14372 | ~11k | ~11.5k |
Inventory info: S/N of F3 BOSEM is 296, F2 301.
Travis and I have taken the 6 DOF transfer function measurements of TMSX. While nothing says it is rubbing (good), there still seems to be something wrong with the top 2 BOSEMs (LF and RT), see attached. The TFs that utilize these 2 BOSEMs show a DC offset across the entire measurment.
Will continue forensics later. Maybe a setting somewhere? TBC...
The changes in the TF look predominantly just a factor of 4-ish increase in sensitivity, and it makes sense to me.
We know that BOSEM LF and RT sensitivity was much smaller before I and Corey repositioned/replaced these on Apr/25. If you look at my alog and Jeff's comment attached (41675), their physical sensitivity increased by a factor of 2.8 and 3.3 respectively for LF and RT, then Jeff changed the calibration constant for LF from 1.295 to 1.604 (a factor of 1.24) and for RT from 1.146 to 1.113 (a factor of 0.97).
The overall sensitivity of LF and RT increased by 2.8*1.24=3.5 and .97*3.3=3.2, respectively, i.e. roughly the P and V sensitivity should have increased by a factor of ~(3.2+3.5)/2~3.4. And 4/3.4~1.2, that's not a big number.
I assume that the reference in Betsy's plot was taken when the sensitivity was already small.
The Mid Y VEA temperatures are being controlled and are at 68degF, permitting us to power up the PEM system there. Jonathan and myself powered up the IO Chassis, the network fiber-converters and the frontend h1pemmy. Since this was an early RCG3.4/3.0.8 test prior to being powered down, I rebuilt both models against RCG3.2/2.6.34 and restarted them.
The latest CDS Overview MEDM once again shows the h1pemmy models. The EDCU has been "greened" by accepting the two missing h0vaclx channels.
It looks like the h0weathermy code sees the comtrol box at MY, but the Davis weather station is not accessible and may need a power cycle.
EY was transitioned to Laser Hazard. Lock out tag for the ALS laser was removed with permission from Keita. Laser is now powered on.
WP 7751
Both the fast shutter and PZT high voltage power supplies for HAM6 are now powered on. The vacuum pressure interlock is bypassed.
Sheila, Ronaldas
Sheila suggested to investigate bullseye sensor noise contribution to DARM. It is basically a coherence based projection of jitter from the bullseye sensor upstream of the PMC, which was contributing to the DARM noise from 30Hz-1 kHz in this lock stretch (4th of July).
Bullseye sensor readings seem to be coherent with DARM signal on YAW and PIT d.o.f., see 'bullseye_yaw_pit_coherence.jpg'.
This sensor measures jitter that we believe is driven by the turbulent flow of water over the crystals in the high power oscillator. The acoustic or PZT injections in the PSL which can reproduce the jitter peaks seen by sensors downstream of the PMC cannot explain the level of noise coupling that the coherence suggests for jitter sensors before the PMC. We are relying on the observed coherence between this sensor and DARM to estimate the noise contribution.
Nonetheless we tried the following:
1) choose maximum coherence d.o.f. for bullseye sensor (either YAW or PIT)
2) if coherence with DARM is bigger than 0.01 for a frequency bin, include it as a noise
3) then the bullseye noise is simply sqrt(coherence) * DARM
Result can be seen in 'bullseye_sensor_coherence_noise.jpg'. This noise contributes quite a lot for f>100Hz and improves noise budget, see 'compact_nb.jpg' and 'full_nb.jpg' [cyan dots in the plot]. For a comparison, previous noise budget plot can be found here (previous noise budget).
While the result seems encouraging, we need to investigate it for other locks. This coherence projection was done for 4th of July, 2017 (GPS time: 1183190418, duration: 600s), before the earthquake. Note that we expect this noise source to be reduced for commissioning/O3 because the high power oscillator has been replaced with the 70W amplifier.
(Kyle R, Gerardo M)
Today after lunch we removed all aux cart equipment used to pump down BSC1 and BSC3 annulus systems, then we moved to do a gross leak check of disturbed joint conflats, background for the entire procedure remained under 10-09 torr*L/sec.
Helium leak tested the viewports/blanks installed on HAM2 and HAM5, and the new installed ion pump gate valves (IP1 and IP2). Kyle noted a couple of viewports that had tape indicating that needed testing on BSC2, leak checked those as well, and removed the tape. No gross leaks found.
Note that we had valved-out the YBM MTP so that all of the helium signal would be routed through the Vertex MTP which was backed by the leak detector. Each joint got a minimum of 120 seconds of audible helium flow, 60 seconds at each 180 degree-separated test slot. Also, following this exercise, we re-exposed the YBM MTP and added the XBM MTP. Thus, the combined volumes of the Vertex+YBM+XBM are now pumped by the (3) MTPs.
[Sheila, Jenne]
When Sheila first opened the light pipe, we weren't seeing beam on IMC REFL camera. We went to the table with a card and moved the PZT until we were back on the camera. I had previously set the IMC optics back to the bottom stage OSEM positions from the last IMC lock. This got us flashes. The IMC is now locked, although the alignment is very poor - we'll continue to work on that after the meeting, but should be ready to send beam to HAM 6 this afternoon.
The IMC is locked with a build up slightly lower than what we had before this vent. On April 26th we had 165 counts on MC2_TRANS_SUM with 2 Watts input, we now have 67 counts with 0.85 Watts of input power. Some of the things that we ran into:
Craig and I re-phased the length demodulator as his comment says, and flipped the sign. We then added 219 degrees to the IMC WFS demod phases. We found that the LSC power normalization was not up to date (probably a consequence of using the rotation stage medm rather than the guardian to set the power) the gain of the IMC WFS was very low. We also reset the dark offsets for the MC WFS. SDF screenshot attached.
Keita found that there was angular feedback to M2 on MC2, which I don't think is correct. This was set this way on Feb 17th and accepted in SDF, we set it back to M2 off and accepted that in SDF.
We set the mode cleaner to offline for a jitter measurement using the IMC WFS DC starting at 1:04 UTC. The PSL is in science mode and there is 8.9 Watts injected.
On a small detour, I made some edits to the state generator for the OFFLOAD_ALIGNMENT_MANY state, bug fixes from the guardian re-work. The motivation for this is that the offload MCWFS script doesn't turn off the loops so it misaligns the IMC which the slow MCWFS loops take a while to recover from. I had intended to write a guardian state that uses this generator, however, this generic offload state relies on having the alignment intergators in the suspension top masses, which is not the case for the MCWFS. We also would need to edit the state to deal with the IMC PZT. There were a few bugs to work out with the generic offload script, I added fast_ezca to the list of things that are imported by ISC_GEN_STATES, and removed two variables which were just renamed in the state ie:
def gen_OFFLOAD_ALIGNMENT_MANY(dof,tramp,optics):
class OFFLOAD_ALIGNMENT(GuardState):
request = False
redirect = False
tramp = ramptime
optics = opticList
became
def gen_OFFLOAD_ALIGNMENT_MANY(dof,ramptime,opticList):
class OFFLOAD_ALIGNMENT(GuardState):
request = False
redirect = False
Craig, Sheila Sheila and I added 4.5 nanoseconds of delay to the IMC PDH error signal. The PDH modulation frequency is 24.078 MHz. This corresponds to a +39 degree phase shift in the PDH error signal.
Attached are data taken with the IMC unlocked with 8.5Watts input power, compared to the data attached to 39434 which was taken with the HPO and the IMC unlocked. Both are calibrated into beam diameters (normalized pit and yaw *sqrt(pi/8)). The 70W amplifier has lower jitter below 50Hz, but not above. This might be due to the high noise on the PSL table and periscope and the floor that Terra and Anamaria have pointed out (41707 and comments).
I did a quick summary page look at the HAM1 ground motion and PSL table motion from November jitter data timeframe and today. I also compared with LLO (used PSL table data from before install started, so August). Floor motion at LHO is slightly lower in the hundreds of Hz region now compared to Nov.
first plot: HAM1 ground, LHO on left, LLO on right
second plot: PSL table motion, LHO on left, LLO on right
