Most of this work was done in September. I thought this stuff was going into one über PEM injection report, but I'll post it here for now. Pieces of this have already been posted in other locations, where relevant. Summary: the only frequency where we have observed coupling from external RF to the interferometer is 45MHz, where the SNR would need to be over 100 in the PEM channel before it shows up in darm.
----------------
We injected RF at specific frequencies where coupling to the interferometer was thought possible. The injections were done using dipole antennas and RF amplifier located by the water tank on site (site chosen for easy access and because its 'far away'). For those injections the 9 MHz and 45 MHz injections were picked up by the PEM radio channels in the corner station electronics bay ~1000x louder than the background. Of the frequencies chosen only the 45MHz injection showed up in the gravitational wave channel.
For the 45MHz injection we saw a PEM antenna SNR of ~1000 and a small peak in darm with an SNR of ~4 when viewed at a ~0.1 Hz resolution. From this we can say that an SNR of 200 in the auxiliary channel would be necessary to produce an SNR of 1 in DARM.
The frequencies chosen are posted at the bottom of the report (in most cases we performed the actual injection at the frequency of interest +100Hz, then looked in DARM for a beat frequency at 100Hz).
Figure 1 shows the 45Mhz injection that appeared in darm. Figures 2 and 3 show a later followup from Oct 23 with a better oscillator source but similar results (at 45 and 9 MHz). The final attachment (pdf) has a summary with more plots.
Future work will include repeat injections with a higher power amplifier and more stable oscillator source.
RF Frequency List (MHz)
9.100230
10.000000
21.500000
24.078360
35.500000
45.501150
71.000001
79.200000
80.000000
TITLE: Nov 6 DAY Shift 16:00-00:00UTC (08:00-04:00 PDT), all times posted in UTC
STATE Of H1: Observing
SUPPORT: Keita, Cheryl, Vern, Richard
LOCK DURATION: ≈2 hrs (this shift)
INCOMING OPERATOR: Jeff B
END-OF-SHIFT SUMMARY: IFO Locked and Observing 78Mpc.
Handing off to Jeff B.
ACTIVITY LOG:
16:00 Peter and Richard going into H1 PSL to troubleshoot RF45 problem
16:05 45MHz issue seems to have corrected itself upon my arrival. I adjusted the RF out gain back to original value as noted in Nutsinee’s aLog(LSC-MOD_RF45_AM_RFSET) aLog #23222
16:06 6.2Mag quake in Aleutians has eq bands up to 10µm/s. Terramon lagged this information ≈33 minutes.
16:15 Joe and Chris headed down X-Arm for beam tube work
17:31 Kyle headed out to X-2-8 to drop off some hardware and then returning.
17:48 Bubba back from EY but heading back for a quick round trip.
18:00 Kyle back
19:19 Keita and Vern will be going into LVEA to further investigate 45Mhz trouble.
20:07 Joe and Chris back from X-arm for lunch
21:12 Keita tell me that I can have the IFO in about 5 minutes!
21:16 Joe and Chris going back out to X-Arm
21:35 begin intial alignment
21:50 Kyle headed back to x-2-8
22:39 IFO locked at NLN
22:48 Robert into LVEA to do injections (≈20min) Then we will proceed to Science Mode
23:34 Joe and Chris back from working on x-arm
2:59UTC
Snapshot before I accepted changes is attached.
I have created a more detailed CDS overview screen customized for O1 operations. It shows the same data as the non-detailed screen and additionally:
SDF data (number of differences and reference file name)
Test Points (number of testpoints opened on the model)
If the SDF diff number is not zero, a red circle is shown. If the numer of testpoints is not zero a blue circle is shown. Note: calibration models are permitted 2 TPs before the circle is shown.
Non-zero CRC accumulated numbers are highlighted with a blue rectangle. A button has been added which permits the operator to clear the accumulated CRC counters.
The screen has been attached to the SITEMAP under the "O-1" pull down.
note that the "DAQ Clear Accumulated CRC" button communicates with the DAQ data concentrator. It would be preferable to only press this when H1 is not in observation mode. The same goes for the command button next to the EDCU red/green status (labled with a single exclamation point) which lists non-connected EPICS channels.
Tagging OpsInfo
MID-SHIFT SUMMARY: The ongoing struggle for 45MhzRFAM stability is the order of the day so far. We’ve been down from a strong Alaskan earthquake since early this morning. The mode cleaner was successfully relocked after a 2µ radian pitch adjustment was made to MC2. There were no ISI nor HEPI trips. I’m awaiting the “go-ahead” to start the re-locking process upon Keita’s approval.
With the control room watching the glitch monitor for the 45MHz. I went through the connections of the RF chain. It seems that the most likely cable the could cause a problem is the run from the Rack by the PSL to the EOM driver in the enclosure. I was able to produce glitches while moving this cable. We have replaced the Cables and the 1db attenuator for this run. This new cable also replaces the short loop used for phasing at the rack end. Hopefully this will fix our problem. If it does not the next move will be the driver itself.
It still glitches once in a while.
I'll install a coupler and a mixer to demodulate the PSL-45MHz signal by the 45MHz distribution amplifier output and monitor it using one of the ADC channels.
Done, nothing was fixed, need to watch out for the next episode (Vern, Keita).
In the attached left, at the beginning after Richard and Fil swapped the cable, RF45 was still glitching. After 19:03:50 mark there was a large glitch, that's when I decided to go in, and I wrote the above alog. The hope was to see if the glitch is likely to come from the upstream or not.
Of course, right after the glitch, it stopped glitching but I didn't know as I was preparing for the incursion.
I and Vern inserted a 20dB coupler to the PSL-45MHz line on the balun on the ISC rack, demodulated it using 45MHz distribution amplifier output from the ISC rack next to it, used a 1.6MHz(?) LPF, a 50 Ohm terminator, and finally SR560 to condition the signal.
SR560 setting was DC in, output zero at 1Hz and pole at 10kHz, with the gain of 2E4 with high dynamic reserve mode. (When DC coupled, the output was 2.3V with DC gain of 200).
The output of SR560 was connected to H1:LSC-EXTRA_AI_1 channel.
We waited for 40 minutes for a glitch but it did not happen (attached, middle and right).
Coupler was disconnected from the ISC rack, 45MHz distribution amplifier output was terminated, but the coupler/mixer/SR560 are left on the floor so it's easily put back on next time.
45MHz phase adjustment
Since RF cable was swapped (and the coil removed), I measured the RF phasing again. Using free swing MICH and measuring the TF from H1:LSC-ASAIR_A_RF45_I_ERR to Q, the phase was measured to be atan(3.45)=73.8deg.
I added a female-female and male-male N barrel in series and it was atan(4.50+-0.04)=77.5+-0.1 deg.
We are shooting for 76.4deg, and last time we adjusted it we ended up 77.3+-0.03, so I decided to go with 77.5+-0.1 deg.
IFO is currently down due earthquake. IMC/FSS stability is still erratic.
Not 100% sure but it seems to me that after this morning's activities, the reference cavity reflected spot (as monitored on video3 in the Control Room) has moved from something close to 12 o'clock to around about 2 o'clock. The reference cavity transmission seems to have increased a little. One can clearly see the effect of the HEPA fan units.
A bit tardy with this:
TITLE: Nov 9 DAY Shift 16:00-23:00UTC (08:00-04:00 PDT), all times posted in UTC
STATE Of H1: Not Good: Environmental/Other
OUTGOING OPERATOR: Travis
QUICK SUMMARY: IFO is locked and the preceding RF45 woes had the range down at 10 Mpc. I was on the phone to the PSL during the meeting. The Short: Kyle is attaching an ion pump to the flange at MY and will start the diesel generator tonight (≈ 00:00UTC); Richard and Peter are troubleshooting and are replacing the cable from the ISC rack. CDS meeting at 22:00UTC.
O1 days 49-52. No restarts reported over all four days.
model restarts logged for Sun 08/Nov/2015
model restarts logged for Sat 07/Nov/2015
model restarts logged for Fri 06/Nov/2015
model restarts logged for Thu 05/Nov/2015
The past 4 days of temperatures in the PSL Laser Room, and the LVEA as seen by the PSL temperature sensors. Nothing appears to be out of the ordinary.
Title: 11/9 Owl Shift 7:00-15:00 UTC (0:00-8:00 PST). All times in UTC.
State of H1: Locked in NLN, but not in Observing due to RF45 glitching
Shift Summary: RF45 glitching does not appear to be improving. Rather than sit in the chair all night collecting bad data, I'm calling it a night and heading home.
Incoming operator: Ed
Activity log:
None, staring at glitchy data.
Keita suggested tapping on the EOM box and wiggling the cable alog 23159. In the absence of anything better, that seems like a good idea.
Tidal and ASC (PRCL1, DHARD) was running away.
The ifo is locked at NLN but not observing. RF45 still glitching.
Lockloss again at 07:00 UTC
DHARD ran away again. I lowered the gain as soon as I noticed the oscillation. The amplitude of the oscillation didn't seem to decrease. Maybe I didn't catch it quick enough?
I also couldn't connect to the nds sever. So no dataviewer.
I will be taking a few minutes to grief.....
Dataviewer seems to work on a different computer. Resume relocking.
Hmmm, maybe we just should always be using the lower DHARD gains when we're on the 45mHz ISI blends? I think DHARD Pit is usually 10, and should go to 7. DHARD YAW is usually 15, and should go to 10.
I don't think that the oscillation that nutsinnee shows in the screen shot is the one that can be fixed by lowering the gains, that is around 0.6Hz, while the screenshots show an something happening at around 20mHz. This large low frequency motion of DHARD yaw was happening durring earthquakes durring ER8.
I would be more inclinnded to leave DHARD gains alone or try to increase the low frequency gain for Yaw, depending on what was really happening (was this ground motion or a loop oscillation?). I think that lowering the gain could actually make things worse in either case.
For a rough model of DHARD yaw loop with thee new boost added, see https://alog.ligo-wa.caltech.edu/aLOG/uploads/21768_20150921225603_DHARDYAWOLGwithboost.png