The LVEA has transitioned to LASER SAFE . This is to facilitate bolting on the HAM 6 doors. Work permit 7565 is complete. This is under work permit 7566.
The SUS_CUST_OPOS_OVERVIEW.adl medm was displaying the damping INMON and OUTPUT in decimal, with 3 decimal places, and I switched those displays to truncated, which matches all other suspension medm screens. When committing the change to svn, I inadvertantly also committed the file SUS_CUST_OPOS_M1_CENTERING_V2ZOOM.adl.
Tvo, Alexei, Dan
We took some more beam profiles of the single bounce (off of ITMX) IFO beam in HAM6 as the north door is off. We took measurements from the edge of the HAM6 table near the viewport and up to OM1. At first glance it seems the astigmatism of the IFO beam is not as bad as our previous measurements. However we still need to analyze the results. We don't have the OPO beam available at the moment so we couldn't profile that, perhaps we can before the doors go back on.
As we were there we also tried to profile the beam shape change induced by the SR3 heater. We got the heater up to ~70C and took profiles with the nanoscan in a single position near the HAM5 viewport. Along with the OMC mode scan results we took the other day we are hoping this will allow us to narrow down where in the beam parameter space we are. Results to come.
Attached is the analysis of the beam profiles taken 50.5'' from OM1 as the SR3 heater temperature increased from 30 to 75 C at 5 W requested power on the SR3 heater.
The starting beam size was (2401.2, 2375.6) um and the final size was (2310.5, 2258.7) um.
The ellipticity (1-b/a) of the beam varied from 0.01 to 0.02.
Attached is the analysis of the last set of data taken during this measurement.
Here the nanoscan was positioned 21' away from OM2, and the SR3 heater was left for about 15 minutes at 75 C to come to equilibrium.
The power to the SR3 heater was then cut and the beam size was tracked as the temperature decayed.
The aim of this exercise was to observe the optical transient of the SR3 by trying to find the point where the beam size would start changing after the power to the SR3 heater was cut.
20 minutes after the power was cut the beam size had not changed significantly and we stopped taking measurements.
Aidan requested some pictures of the EY HWS setup on the table in preparation for next weeks upgrade, pictures attached
For reference, the layout is given in https://dcc.ligo.org/D1400241
For reference, the layout is given in https://dcc.ligo.org/D1400241
Apollo TJ Daniel
All squeezer beams are clearing the viewports.
The south door is tacked on and can be torqued down.
SQZT6 has been moved into its final position and its feet outlined on the floor.
A note on the process: to make sure that the beam was actually going to make it out, Jenne and I first placed the viewport emulator on top of the HAM6 N door that was laid down flat. Since the emulator was designed for iLIGO viewports, the extra covers and shrouds and other safety features make adjusting the emulator rings basically impossible. We didn't adjust anything because it looked to be good enough for what we could tell here. After Dan B, Jenne and I rearranged the flange protectors and then mounted the emulator, I pulled out the plumb line and the center emulator ring, the main one we were worried about, seemed to be in the correct spot. Daniel double checked with me and we agreed that it may be off slightly in the horizontal, but we could work with what we had.
Now for the aligning. I had a laminated piece of paper with a hole the diameter of the viewport cut out of it that we could place on the ring to make sure that it would actually go out (see attached). We also marked the wall with a dry erase marker to easily gauge how much we need to move the beam (see attached). We got them to a good point and then checked the the beam diverter beam was good as well, it was. With the door tacked on we did a final check it it all looked good (see attached).
[DanB, Jenne, TVo, Alexei, Daniel]
Upon heading into HAM6 today, we found that the AS_C beam was quite far off the QPD, hitting the case. This required about 3 times as many picomotor steps to get the beam onto the diode than I was using yesterday during my raster scan. When we got in there, the yaw actuator was near the edge of its range, but recentering the beam brought it back closer to the center.
We were a bit confused later in the day, that it looked like the picomotor wasn't actuating in the way that one would expect, but I think it turns out that I had moved the beam off the QPD and not realized it. When we went back in after lunch and redid the centering of the beam, everything looks normal again (as it did before lunch).
We see PRMI and MICH fringes on the diode with 5W injected into vacuum, in single-bounce config. The QPD also passes a flashlight test - since the diode's back faces the doorway, we used a clean dental mirror to reflect a flashlight beam onto the QPD, and we see signals on all 4 segments. With the beam confirmed on the QPD when flashing, we misaligned ITMY and PRM, and then are able to steer the beam around the diode using the picomotors.
So, in the end, nothing was broken, just super misaligned. We likely would have found this, if we had done a much larger raster scan than I did yesterday. I'm glad we know this now, at a time we were pulling the door anyway, rather than having to vent specifically for this.
TVo, Dan, and Alexei are going to take some beam scans of the main beam today and tomorrow morning as part of the diagnostic of the astigmatism problem we think we have, but we're likely ready for the HAM6 north door to go back on tomorrow or Friday. We'll decide at the morning meeting tomorrow.
The data retransmission rate between the data concentrator and the frame writer machines has been higher than we would like as we have upgrade the OS of the daqd systems to Debian 8. WP 7558 was opened to explore changes on the spare/test frame writer to see if we could lower the retransmit rate. After looking at timings yesterday one of the 'producer' threads, which verifies checksums of the data as it comes off the network, was found to be running long at times. Today I put in more efficient checksum code* into the frame writer. This gives us an immediate savings of 15ms and a more stable timing of the thread in question. In the attached image you can see a few important points. 1. The PRDCR_CRC_TIME_CRC_MAX_MS value is 15ms lower. 2. Thus the total time PRDCR_CRC_TIME_FULL_MAX_MS drops as well increasing our safety margin 3. This is a big improvement on the spike at the 4m mark which jumps above the allotted time for the thread (this spike is ~70ms!). * This was code that John Z. had done for the nds2 server. Note for Dave Barker: this is a new daqd that has a new framecpp version string in the full frames, so the checksums are not expected to match between the full frames for h1fw[01] & h1fw2.
We let h1fw2 run overnight. I've attached a plot of the number of re-transmit requests over the last 44 hours (from around the last daq restart). The left side is the old code, the right side is the new. We are still seeing the re-transmits, but the rate is looking better.
Slawek, Sebastien, Travis
After Slawek and Sebastien's whirlwind trip over from the east coast last night and this morning, they joined Travis to glue AMD #1 and #2 onto the flats of the ETMX test mass in-chamber. They once again encountered no issues with executing the procedure or using the tooling. Well done, team.
#3 and #4 will be glued onto the flats tomorrow late afternoon.
The LVEA has transitioned to LASER HAZARD Work permit 7564 is complete. This is under work permit 7565.
I took snapshots of the IO GigE camera 1 image today with the gain set to 40000 and 45000. Attached is a snapshot of the camera images, and cross sections from top to bottom and right to left.
I've been running the camera at a gain of 50000, but this saturates the center of the beam. The new camera gain is set to 40000, where no pixels saturate.
Further investigation into the vault electronics was conducted today. We hit the reset on the seismometer re-centering it, Robert says it looks "ok" now.
We continued work on the LEMI magnetometers. X-LEMI is not functioning at all, Y-LEMI functions in all configurations. It looks as if the X-LEMI itself or the cable leading to it is not working. We adjusted the LEMI power input from +/-6V to +/-7V at the supply but this did not help the X-LEMI. These LEMI devices are designed to run on car batteries, from +/- 12V down to +/- 6V.
The LVEA has transitioned to LASER SAFE Work permit 7562 is complete. This is under work permit 7564.
Ed, Jason
Peter K. (WP #7562) transitioned the LVEA to laser hazard.
Alexei, TVo
When we did the OMC scan while varying the SR3 RoC power, we noticed that the power monitor which is the digital product of an analog voltage and current monitor (found here) would have very odd behavior. When we would request between 0.0 and 0.8 Watts, the current and voltage monitors would respond with the correct values we would expect from the simple equations (P=V^2/R=I^2*R). However, when requesting above .8 Watts, the voltage monitor would rail and we'd get some nonsensical values. We double checked that our drive calibration was correct in analog land, and everything turned out as we'd expect.
We narrowed the problem down to what I think is a design issue, the voltage monitor when requesting say 0.8 Watts reads 9.7 Volts (positive-negative) at the input of the ADC. The EL3104 modules only has a range of +/-10 Volts which means than any higher requested power will rail the ADC and gives back a nonsensical voltage reading. I'm not sure how Beckhoff likes to throw and error if this is true. If I'm correct, we need to either switch to an ADC with a larger voltage range or an analog filter that attenuates the input signal so that we can get the full range of what we expect when energizing the heater.
I measured the AC coupling loop OLTF of the ISS 2nd loop with the same setting as O2 and the phase margin was less than 3 degrees (first attachment).
I had to flip the sign of the loop, as the whitening in ERR1 which is used for this loop was changed from two inverting opamps (p=2.7k and 130kHz in V2) to just one inverting opamp (p=2.7k in V3) in D1600298. Our phase margin might have been this thin during O2 since the loop apparently works and the AC coupling is only necessary during power ramping.
Anyway, this is a bit too extreme, for the moment I'll decrease the gain by ~10dB (H1:PSL-ISS_SECONDLOOP_AC_COUPLING_SERVO_GAIN=-0.08 instead of -0.25, second attachment). If the IMC cannot take the power ramping with this setting we'll redesign the filters.
(Ignore 0.1Hz^2 Sallen-Key low pass in the AC coupling path on the medm screen, that filter is off, it's a logic error in the frontend model which I'll fix.)
(The measurement was done without light since we don't need it and since I didn't know the status of ISS 1st loop at the moment. According to Jason AOM isn't installed yet.)
