Sheila, Kiwamu
This morning I put in upgrades for both guardian core (r1076) and cdsutils (r322). All guardian nodes have been restarted (except for ITMX SUS and SEI which are under test by Jim W. Will restart as soon as he's done).
cdsutils r322 new improvements/features:
guardian r1076 improvements/features (see also LLO aLOG 13930):
Here's the updated guardian control screen that comes with this upgrade:
Note the new "NOMINAL" state field. The NOMINAL state is NONE by default if not set in the code module (which is the case here for SUS_BS). The orange background reflects the fact that the new OK bit is False, which in this case is because there is no NOMINAL state set.
Also note the number fields to the right of the state names, which show the numeric index for the current states. Now that we manually set these numbers, they have a bit more meaning, which is why it's useful to display them.
We now need to go through all systems and add appropriate state indices and nominal state definitions. This was done already at LLO, so I'll be adopting the same standards.
ITMX SUS and SEI nodes have now been restarted, although the SEI is currently tripped as Jim W. is still working on tuning their loops.
J. Kissel Picking up where Borja left off, I've measured the charge H1 SUS ETMY test mass / reaction mass system. Today's measurement is made in hopes of seeing if the ion pump is the smoking gun causing all the charging noise. Driving quadrant-by-quadrant, and using the optical lever's Pitch and Yaw signals as my response, with a 0.02 [Hz] binWidth, with 5 (instead of 3) ensuring I had a coherence of at least 0.75 (i.e. an relative amplitude uncertainty of less than sqrt((1 - 0.75)/2*5*0.75) = 18%), and taking 5 bias data points (instead of 4), I arrive at the following effective bias voltages: P [V] Y [V] UL 104 149 LL 129 72 UR 110 13.6 LR 104 117 For some reason Pitch had much better SNR for the same drive amplitude, but given that I accounted for the coherence, both pitch and yaw are to be believed. I'm not sure, however, why pitch shows such a uniform charging distribution, where yaw reports the charge is spread unevenly. I'll gather more measurements and plot the results against Borja's previous measurements tomorrow. Since we expect this to evolve with time, there's not to much point in putting it up against the ion pump closed results just yet. I've got some clean up functions and scripts to do, bear with me while I catch up to what Borja had assembled. I could have ask him for his analysis scripts, but it was easier just to write my own, given that the analysis is so simple. Still no automation of this measurement, and given the first-day-back barrage of Q&A, it took me ~4 hours to complete the measurement. I expect this will go faster once I gather my stride. As usual, it's a toss up between "let's just get the answer now" and "take the time to debug the automation." I elected for the former.
Lisa, Sheila
We went down to end X, turned on the laser, adjusted the fiber polarization, adjusted the laser temperature to find the beat note, tuned up the beat note alignment, and the PLL is now locked. The QPD servos are also locked.
There is 6.5 mW on the PLL BBPD, the medm screen reports -5.6 dBm beat note strength, and the laser temperature is 28.96 C.
There's a relatively large PIT motion visible in OMC_REFL and AS_AIR video.
If you look at OMC QPDs, it turns out to be a broad peak at around 0.7Hz, and is mostly coherent with SR3 oplev and nothing else.
I did an svn update on the main SUS guardian module, sus/common/guardian/SUS.py, now at r8553. This pulled in recent changes from L1, including the addion of state indices and a NOMINAL state definition. I commented out some new additional OPLEV logic that we're not ready for here yet. All SUS nodes were RELOADED for these changes to take affect.
Here's the new guardian control screen for SUS_BS with the new changes in place:
Notice:
The updates also included the addition of some optical lever (oplev) damping loop triggering, that I have commented out for the time being, until we get the oplevs working here (and the logic better).
In the event of a power loss during a science run, the APC Smart-UPS (Model: 1500, Max Configurable Power: 980 Watts/1440 VA, Mfg. Part: SUA1500R2X122) will continue powering the Pre-Stabilized Laser (PSL) to prevent laser damage. The UPS (Uninterruptible Power Supply) draws for its transformer a large amount of current, so we measured the device's magnetic field at various distances to determine the 'minimum distance' -- how close the UPS can be to other equipment, especially the interferometer. We defined the maximum allowable magnetic field at 60 Hz from any electronic device to be 0.4 nT, which is one tenth of the root mean square 60 Hz magnetic field during iLIGO science runs. For magnetic field measurements, we used two 500 Watt lights (to simulate the PSL load), a Bartington magnetometer (mounted on a tripod), and the UPS (placed horizontally and face-up on a plastic bin). During preliminary tests, we found that the system's on/plugged-in and off/plugged-in configurations produced similar magnetic field magnitudes. Therefore, the UPS must be placed at its minimum distance whenever the device is plugged-in (and either on or off). Next, at 1m, we measured the magnetic field at three different angles relative to the physical center of the device. Of these measurements, The lowest value was 55% of the highest value. The magnetic field at 60 Hz was strongest when the magnetometer was aligned with the device's front (face with buttons) left edge, and the remaining measurements were taken at this angle. The final step in determining the minimum distance was measuring the attenuation of the device's magnetic field with distance. To minimize noise caused by surrounding electronic devices, I collected data in the VEA of End X instead of in the LVEA. I took data from 3 to 39 feet, with a constant interval of 3 feet. At each distance, I used Diagnostic Test Tools (DTT) to record the power spectra of the magnetic field when the UPS was unplugged and off and when the UPS was plugged-in and on. I took two data sets--both were numerically similar. The data was slightly noisy, so I used Grace to perform a linear regression on, at 60 Hz, the natural log of the magnetic field vs. the natural log of the distance. I found the fitted curve to be described by: y=(170.579)x^-1.739 In conclusion, when y is 0.4nT (see above), x is about 33 ft -- the minimum distance. The attached plot shows the UPS magnetic field attenuation. Here is a link to the UPS at the LIGO Hanford Observatory: http://www.apc.com/resource/include/techspec_index.cfm?base_sku=SUA1500R2X122 Christina Daniel, Robert Schofield
I did some math to figure out how much ITMX, ITMY and BS may have been moving (in a frequency band of 0.1-1 ish Hz) in their angles according to fluctuation of the DC light observed at the dark port when the Michelson was locked.
(Summary)
Wednesday night (August 27th)
Friday night (August 29th)
(Some math behind it)
Suppose the Michelson is locked on a dark fringe. If an ITM is misaligned by Ψ, this introduces a displacement and tilt in the reflected beam with respect to the one from the other ITM at the BS. The displacement is x = 2 L Ψ and the title is ϑ = 2 Ψ where L is the distance from the BS to the ITM. So we get a small amount of 01 or 10 mode at the dark port on top of the 00 modes. Since the effect on the resultant 00 mode in its power is proportial to 4th power of the displacement and tilt, we assume the 00 mode to vanish because of the locking loop. The only residual we obtain at the dark port is the 01 or 10 mode whose field can be written as
E10 = 1/sqrt(2) ( x/w0 + i * ϑ / ϑ0),
where w0 is the waist size and ϑ0 is the divergence angle respectively. A factor of 1/sqrt(2) upfront comes from the BS reflection. If we plug the definition of x and ϑ into the equation, we get
E10 = sqrt(2) ( L/w0 + i / ϑ0) Ψ.
Squaring the above, one can get the dark port power as
P = 2 ( (L/w0)2+ (1/ ϑ0)2) Ψ2
Note that P is already normalized by the input beam power or equivallently the bright fringe. The Rayleigh range of the beam around the BS is roughly 210 m (if my math is correct). This gives a waist size of 8.4 mm and divergence angle of 40 urad. The ITM-BS distance L is about 5.34 m where I averaged out the Schnupp asymmetry. So the dark port power can be now explicitly written as
P = 1.24 x 109 Ψ2
This is the equation I used for deriving the numbers listed at the very top.
For example, if one wants to explain a 16% DC light fluctuation observed at the dark port by an angle deviation in ITMX(Y), the misalignment should be Ψ = sqrt(0.16 / 1.24 x 109 ) = 11.4 urad. In the case of the BS, the effect gets twice bigger due to the fact it affects both X and Y beams at the same time in a constructive manner.
Nic, Lisa Somehow I was confused by Kiwamu's final numbers, so we went through the math again. Kiwamu is correct. The 11 urad seemed huge for a 16% power fluctuation, but Kiwamu is referring to 16% power fluctuations with respect to the BRIGHT fringe..so it is indeed huge.
Patrick, Jim, Dave
we reconfigured the conlog to get the unmonitored number from 3000+ down to zero. We moved the susquadtst targets into the target archive. We had to hand edit the pv_list.txt to replace H1:ODC-CAL_PCALX_* channels with H1:CAL-PCALX_*. There is a bug in my include list generator python script which did not handle the top-names correctly, needs a rewrite using the autoBurt.req files instead.
on opsws1. Please don't disturb the matlab session. Except for watchdog trips, all is automatic.
The X-end controller for the baffle diodes was still one of the early units with parts missing (BO at the time). Swapped S1302233 with S1400075; and updated the Beckhoff to reflect these changes. It is now possible to adjust the gains remotely.
It seems like the damping loops are OK for the SRs. Attached are screenshots comparing each of them to the corresponding PRC optic. I didn't adjust anything on SR2 or SR3, on SRM I turned off FM1 on L, to match the configuration of the PRM damping loops. I also increased the pitch gain from -2 to -3, since there was more pitch motion on the SRM, the resulting spectra are in the screenshots below.
SR3 T+V have extra noise at around 13.6 Hz and 16.25 Hz respectively. This is there with damping on and off, and there are no resonances predicted by the model at these frequencies. Jeff suggests that these could be HEPI resonances.
To do this, We had to turn on some ISI loops on HAM5. To do that, Jamie had to edit the guardian so that now we can request ISOLATED_ROBUST. Hugh also had to reset the CPS targets.
Based on the top stage osems, it seems like the motion of the SR cavity should be comparable to the motion of the PRC, so large motion of the optics shouldn't be an impediment to locking the SRC anymore now that HAM5 has some form of isolation.
LVEA is Laser Hazard 08:29 Hugh - Going to End-X to check HEIP pump controller and put H2 actuator into run mode 09:14 Filiberto – In LVEA working on cable cleanup around the spools 09:36 Richard – In LVEA 09:45 Praxair Nitrogen delivery 09:47 Jeff K. - Going to End-Y to check the ION pump status 09:56 Peter & Rick – Going into the H1 PLS enclosure 10:07 Richard – Going to End-X 10:10 Jodi – Going to End-X and then to End-Y to check property tags 10:17 Hugh – Going to End-X 10:35 Betsy & Travis – Going into the LVEA to look for parts 10:35 Kiwamu & Lisa – Checking out electronics at ITC6 11:10 Unifirst on site to change out entryway mats 11:26 Bottled water delivery to site 11:50 Check DR chiller water level. Level is good 12:29 Filiberto – Checking for ground loops on SRM 13:00 Cris – At End-X 13:03 Karen – At End-Y 14:40 Wire delivery for Richard
With the new Parker Vlave on H2, ran Range of Motion and Linearity tests on the system. Before I could do that I saw that a zeroing of the IPS was in order. Some sensors were too far out to really test the range of the Actuator. Rezeroing the IPS will change the Cartesian Basis Positions and therefore the loop targets. I trended the Positions back to IFOX and then applied the free floating to Isolated deltas to the new near-zero positions for RY & RZ. Remember too that EndX had the large Ry tilt for the TMS pointing. Keita turned off the pitch loop and then adjusted the TMS accordingly. Anyway, straight forward shift it was that and is now this:
Pos |
Target |
|
X | -6100 | -6700 |
Y | -10000 | -10600 |
Z | -14300 | -15200 |
RX | -9100 | -9100 |
RY | -2900 | -1800 |
RZ | -600 | 7300 |
Guardian only holds the RZ and RY targets(at EndX;) the other DoFs are locked to the free hang position at the time of Isolation start so will drift around with the usual HEPI hysteresis. The other targets I just grabbed and are really only for reference. If the free hanging position is terribly far from these numbers (no, I don't know what too far is..) something may be amiss.
Back to the IPS zeroing, the mechanical stops on the HEPI Actuators (not adjustable) were mostly compatable with the IPS zeroing. ROM was 1.0mm for all Actuators except H2+, H4+ & V1+; H2+ & V1+ managed 0.9mm, and H4+ made 0.8mm. Linearity test results are attached; Jan results in first plot, today's in second. The new valve does seem to have less umph (from a new batch hot from LLO calibration) dropping a little less than 20% compared to the . Still, the position loops close and are stable. I have some other evaluations to do and will update.
safe.snap updated for ETMX HEPI.
The new H1 ITMs ROC (ITM03 and ITM11) are similar to the ones in L1, but they are swapped (the wavefront error is larger from X than from Y). Based on T1300954 (table 3) and Hiro's wisdom, the effective ROCs of the H1 optics, as measured in reflection, going through the bulk, are: R_ITMX (ITM03) = 1939.3 + (-10.92*2*1.457); R_ITMY (ITM11)= 1939.2 + (1.56*2*1.457); By looking at the L1 data in single bounce without TCS (below), one should expect something like ~20% mode mismatch for X and something somehow better for Y. L1 Mode mis-match: NO TCS: ITMX 14.5% ITMY 22% Even with an input beam perfectly matched to the PRM, I would expect something like: modematching asX with OMC = 0.8408 modematching asY with OMC = 0.91229
To improve the contrast while maximize the matching to the OMC, CO2 central heating should be applied to ITMX to match ITMY. Since we don't have central heating right now, one could use the ring heater to match ITMY to ITMX. This would make the matching to the OMC worse, but a better contrast.
See 13815 entry instead.
Reset WD counters on HAM5, ITMX, BS, and ITMY.
Nic, Kiwamu
(Variable finesse technique turned out to be good to start)
Since I have been unsuccessful in locking the PRMI in the past two or three days, I wanted to try some other locking technique. We tried LLO's variable finesse technique (see LLO alog 11340) which seemed more reliable than randomly adjusting the gains and triggers. It turned out that it almost repeatably locks the initial low finesse PRMI. Very nice. We then fiddled with the MICH gain which needed some gain correction as we got rid of the offset in the MICH locking point.
POPAIR_B_LF fluctuated a lot presumably due to some misalignment in some optics. POPAIR_B_LF was about 20000 counts in average and ASAIR_B_LF stayed approximately 3000 counts in average. After 15 minutes or so, we lost the lock for some reason, we did not have a close look.
The attached is a video of ASAIR when the PRMI was in lock.
The final configuration (i.e. the MICH locked on a zero-offset point) is shown in the attached screenshot.
Also, we newly installed a 100 Hz low pass filter in POPAIR_B_LF because high frequency noise in POPAIR_B_LF saturated the BS actuator through the normalization.
When the simple Michelson is locked on a dark fringe, POPAIR_B_LF is typically 120 counts. So the recycling gain is (20000 counts) x (Tp 3%) / (120 counts) = 5 which seems too small. Clipping loss somewhere ?
For the records, with 60W in the L1 recycling cavity, without the BS baffles the BS drift in PRMI carrier lock was about 5 urad (see LLO entry 9920). This is the only H1 PRMI carrier lock collected so far, but the only drifts we see are ~0.5urad.