Summary: 1) High magnetic coupling at HAM5 supports checking the magnet orientation on the bottom stage of SRM and SR3. 2) 600-700 Hz coupling at BS is most consistent with ST0 source, like the elliptical baffles. 3) There appears to be scattering at the P-Cal ports that is not yet addressed. 4) Cell phone-band injections were not seen in DARM. 5) Likely scattering noise associated with shaking the BSC2 chamber may be limiting.
Because of the vent schedule, less time than requested was allotted for follow-up PEM injections. The two evenings available were partitioned so that all topics were addressed, but not in great depth.
1) The high magnetic coupling in the output arm is at HAM5. The highest magnetic coupling at LHO, measured during PEM injections, was in the output arm (Figure 1). The coupling may be limiting at 14 Hz and below, and would dominate the coupling of Schumann resonances. To search for the coupling site, mobile magnetic injections with large and small coils were used while monitoring DARM. The angles of the coils were changed to maximize coupling. The strongest magnetic coupling was found near HAM5; Figure 2 shows an order of magnitude stronger coupling to DARM for an injection next to HAM5 than for a similar injection a couple of meters away and next to HAM6.
A fit of the coupling factors gave a frequency dependence of 1/f^4.65 (R^2=0.998). For length coupling to the magnets on SRM or SR3, we would expect 1/f^3 (1/f^2 for acceleration to displacement and 1/f for eddy current shielding). One possibility is that the coupling is angular. In any case, we should probably check the orientation of magnets on these optics.
2) The coupling for HEPI shaking of the BS is not consistent with ISI suspension resonances. PEM injections indicated high coupling in the 600-700 Hz regions for shaking of the BS table with HEPI (https://dcc.ligo.org/DocDB/0144/G1701613/002/GenevaTalkSm.pdf), There are ISI suspension resonances near these frequencies, but follow-up investigations did not support this hypothesis. For example, the peaks did not show up for injections at stage 2. Instead, results were more consistent with coupling at stage 0, like the elliptical baffles that are also implicated at frequencies between 70 and 200 Hz.
3) There appears to be significant scattering from the P-Cal beam RX port, which may need to be mitigated. I mounted a small shaker on the RX port for the EY P-Cal beam and compared the coupling to that from a larger shaker mounted on the reduction flange (photos in Figure 3). Figure 4 shows that there were peaks that were much larger in DARM for the small shaker on the port than for the global shaker, even though the shaking was approximately balanced at the beam tube around the P-Cal periscope. This suggests that resonances of the RX port are modulating scattering. The RX box is connected via a rubber bellows and is thus isolated from the shaking, so the source of scattering is likely to be the flange containing the window or, possibly, the window itself. I used an accelerometer mounted on the port to estimate the contribution of ambient motion to DARM at 70 Hz: it was a factor of 6.8 below DARM.
Possible mitigation schemes would include adjusting the periscope mirror to eliminate direct retro-reflection of the ETM beam spot, and baffling the flange. The TX port is also a potential remaining scattering site, but I didn’t get a chance to investigate it.
4) Gigahertz injections in the CS electronics bay did not appear in DARM. Cell phone signals travel on line-of-sight paths and so are not a concern for coupling between sites. Furthermore, our experimental areas are fairly good Faraday cages, but cell phones in the building are ever-present potential sources of noise. As a brief preliminary check, I used an SRS SG382 with a Mini-Circuits TVA-11-422A+ amplifier and lambda/2 antennae. I broadcast full blast in the electronics bay with an 800 MHz signal modulated by noise (random on and off every 0.1 microseconds). No evidence of coupling was found in DARM.
5) Shaking BSC2 (BS chamber) suggests that ambient vibration may be limiting at low-f and an ~order of magnitude increase over ambient vibration below 18 Hz produces 100 + Hz signal in DARM. PEM injections with a new shaker demonstrated coupling in the input and output arms (https://dcc.ligo.org/DocDB/0144/G1701613/002/GenevaTalkSm.pdf). The output arm measurements were complicated by coupling of the large magnetic fields produced by the heavy-duty shaker and the high coupling of magnetic fields in the output arm mentioned in "1)" above. In the follow-up PEM injections, the magnetic coupling problem was reduced by coupling the shaker to the vacuum enclosure with 8 foot PVC pipe, allowing the shaker to be further from the magnetic coupling sites. Coupling was evident in the output arm and HAM5 (possibly the HAM5-6 septum which is attached to HAM5). However, there was not enough time to narrow down the frequency bands and the coupling locations. But these sites may be mitigated by improvements in baffling planned for this area. No coupling was detected from shaking the X-manifold.
Figure 5 shows the coupling from broad-band shaking of the BSC2 vacuum wall, as well as the contribution of magnetic coupling from the shaker. Figure 5 also illustrates the shaker setup and shows how the magnetic coupling was estimated by detaching the PVC pipe. Some of the lines in Figure 5, like 70, 100, and 105 Hz, are also produced by HEPI shaking. Shaking of the chamber walls also couples through the floor and shakes the blue piers. The various motion sensors were consistent with the shaker producing the lines by shaking ST0 of the ISI.
The broad band noise evident between 50 and 300 Hz in Figure 5 appears to be produced by motion between 8 and 18 Hz and is a factor of 2 or so above DARM for an order of magnitude or less increase over ambient vibration below 18 Hz (Figure 6). It is likely, but not certain, that this noise is produced by vacuum enclosure motion. The uncertainty is because, between 14-17.5 Hz, the chamber shaking produced greater motion at ISI ST0-ST2 in the X and Y axes than were tested with HEPI shaking. The direct coupling between 10-21 Hz may keep us from reaching 1e-19m/sqrt(Hz) in the 10-21 Hz region.
Preliminary investigations, while I was helping remove the Swiss cheese baffle center, suggested some possible scattering sites. I would like to investigate these further by comparing the resonant frequencies of these sites to the peaks that the shaking produced in DARM. This will require external/internal accelerometers. I would also check if the elliptical baffles have resonances at 70, 100 and 105 Hz.
Robert S, Philippe N., Anamaria E.
