h2susb78, the replaced front end computer for FMY, crashed again at 5:55pm after just over 6 hours operational. So it looks like the problem may be in the IO Chassis or the communications with it.
Attached are plots of dust counts > .5 microns.
Grouting and chamber door close at End-Y BSC 6. Work on BSC 8 in the LVEA
(corey, eric, hugo, mitch)
#7
Started assembling Stiffener under Stage-0. Managed to get about 75% done, and called it an early day due to high temperatures in Staging (swapping HVAC filters took temperature from 65deg to 72). So, we didn't want to have more sweat than normal drop on Class-A parts.
Horizontal L4Cs
Helped Hugo troubleshoot issues with corner 3 L4Cs (we couldn't see their signal). Hugo eventually tracked down the issue to all (3) BIO Interface boxes & Corner (3) Anti-Alias boxes being turned OFF.
On behalf of Bubba, I write:
The cleanroom at BSC6 seemed to be drying the new grout faster than desired, so the cleanroom is being turned off for the weekend.
On the latest shipment of GS-13s to be repaired at LLO we included a G-View shock recorder. The G-View is 25g max and we had set the threshold for recording at 10g. On the graph you can see that it records 4 data points when any of the 3 axis (x y horizontal plane z vertical) 10g threshold is reached. According to LLO the battery had worked loose when it was received (on the 31st) but it doesn't look like too much data was lost as our last data point was the 30th. Hopefully we can gather some more data as we ship containers back and forth.
The welders are moving the plumbing already build at BSC7 to BSC2. They've managed to disconnect without cutting. Could be done with BSCs 2 & 3 pretty quick! See photo for the modular design!
I swapped out many of the 3" shrouds for 4" versions where needed and possible. All cables are strain relieved to either the bolt through the shroud or to the Cable tray where necessary. See the photos below for a typical cluster.
For the record, this is the count of the PEEK cable ties and clamps installed in BSC6 on the SUS, ISI, and TMS:
(~13) D1000600 sets on top side of ISI
(13) D1000600 sets on underside of ISI Stage 0
(6) D1000600 sets on underside of ISI Stage 2
(2) D1000600 sets on the TMS structure
(10) PEEK Zip ties on the ETMy QUAD structure
Each D1000600 set is comprised of 2 of the D1000600 PEEK clamps, so we used a total of
68 D1000600 individual clamps
BSC2 flooring was moved into the LVEA. The crew went into HAM4 and removed the dust barriers. BSC2 flooring was installed. (There was a slight delay caused by a missing nut.) The doors will go back on this afternoon.
h2susb78 remains unstable. It crashed between 2am and 3am this morning. It was removed from the watchdog system yesterday, and did not trip SEI BSC8 this time. I will power cycle it, and it should not crash SUS ITMY since it is no longer part of the Dolphin fabric.
CPU was replaced with another H2 frontend with the same card configuration. New machine was restarted at 11:30 today. h2boot was modified to take the new MAC address.
We relocated one of the two Qlogic fiber channel switches from the H2 EE building (where it was part of a redundant CDS-LDAS network) to its new location in the MSR. It will connect the new H1 DAQ to LDAS, allowing the publishing of H1 frames and distribution of H1 data via the NDS2 server. Dan is setting up the QFS solaris machine for h1fw0 (we currently only have h1fw1 operational). Jim is building up h1fw0 from its cloned state.
While h2susb78 remains unstable, FMY has been removed from the IOP watchdog system. This required a code change with the h2iopseib8 model to disconnect the FMY signal from the IOP SEI BSC8 DACKILL part. All the models on h2seib8 were then restarted which included a new ISI model which also had FMY removed from its watchdog. The frozen h2susb78 was powered down and its Dolphin IPC cable was disconnected before it was powered back. This will ensure that any possible future crashes and subsequent power cycles will not stop the ITMY model.
There was an operator error this afternoon around 18:44 PST where a script being tested for damping was inadvertently started on the ETMY SUS. The script was meant to switch off inputs to the damping filters but caused the user model and IOP watchdogs to trip. The ISI DAC output has been interrupted, but a SUS ETMY measurement is slated to begin tonight.
Attached are plots of dust counts > .5 microns. The dust monitor at location 15 in the LVEA has been turned off.
The day's main activities: End-Y BSC 6 dome cover and door installation. DAQ reboots by CDS Simulink model reboots on the BSC 8 chamber models (FMY, ITMY, SEI) by CDS Removal of FMY from the IOP watchdog monitors
D. Barker, J. Garcia The optical lever photodiode signals were added to the H2 SUS ITMY Simulink model ('h2susitmy.mdl') at the "L3" stage. The signals are routed to the second ADC on 'h2susb478', which is labeled "ADC_1" in the Simulink model. The "h2susitmy.mdl" file was modified to incorporate the four quadrant signals from the ITMY Optical Lever photodiode into the "QUAD_MASTER.mdl" library part. In the 'h2susitmy.mdl' file, the four quadrant signals are on the ADC_1 channels: UL = ADC_1_1 UR = ADC_1_3 LL = ADC_1_2 LR = ADC_1_4 NOTE: this configuration will hold true for the One Arm Test and does NOT reflect the latest H2 SUS ITM, BS, FM wiring diagram, D1001725-v10. ------- In addition to the above changes to the "custom" top level Simulink model ('h2susitmy'), the "QUAD_MASTER.mdl" generic library part was modified for several bug fixes. Inside the 'QUAD_MASTER' diagram, the editing was performed mainly at the 'QUAD_MASTER/QUAD/L3/OPLEVWIT/' level. Initially, SUM, Pitch, and Yaw outputs were test points without filter modules. However, this prevented the DAQ from accepting these channels and writing them to frames, unbeknowst to CDS. These library parts were replaced with actual filter modules instead of test points. This resolved the ability to write these channels to frames. The Optical Lever photodiode channels being written to frames are: H2:SUS-ITMY_L3_OPLEVWIT_P_OUT H2:SUS-ITMY_L3_OPLEVWIT_Y_OUT H2:SUS-ITMY_L3_OPLEVWIT_SUM_OUT The 'h2susitmy' model has now been recompiled and installed on 'h2susb478'.
In addition to the above changes to the Simulink model, there were two other modifications. The inter-computer communication library part (PCI) was removed, because the Optical Lever signals are now fed directly to the H2 SUS ITMY machine 'h2susb478'. Secondly, the input for the Laser Power monitor was grounded. Currently, the laser power is not directly fed to the ADC on 'h2susb478'.
J. Garcia, S. Steplewski We made a set of TOP transfer function measurements on ITMY Tuesday. This was the first time we tested ITMY since this chamber was pumped down. Measurements were clean and the results match up reasonably well with the (fiber) model - although this is a wire rehang after the fiber break. The .mat files are located in: ~/SusSVN/trunk/QUAD/H2/ITMY/SAGM0/Data/ ~/SusSVN/trunk/QUAD/H2/ITMY/SAGR0/Data/ The .pdf files are located in: ~/SusSVN/trunk/QUAD/H2/ITMY/SAGM0/Results ~/SusSVN/trunk/QUAD/H2/ITMY/SAGR0/Results
Attached is another plot that shows the ITMY TOP transfer function taken In-Vacuum compared to that taken In-Air. Aslo plotted are the wire (for M0) and thincp (for R0) models for reference.
Mark Barton, Szymon Steplewski We changed the parameters in the Mathematica model for the QUAD suspension for ITMY to reflect the fact that the glass test mass is suspended from steel wires after the fiber break. These changes have resulted in noticeably better agreement between the measurements and model for several degrees of freedom. The TOP Pitch to Pitch mode around 1.4 Hz has shifted to higher frequency (from about 1.31 Hz to 1.44 Hz), while the Pitch mode near 0.56 Hz is practically unchanged, which is where we want it according to the design requirements. The TOP Longitudinal to Longitudinal plot also shows some nice improvement in the mode around ~1.5 Hz, which matches up a little more with the "wirerehang" model now. There is now a new wirerehang Mathematica model in the SUS SVN under ~/trunk/Common/MathematicaModels/QuadLite2Lateral/mark.barton/20120601TMproductionTMrehang/ which was built from elements of the fiber model under ~/trunk/Common/MathematicaModels/QuadLite2Lateral/mark.barton/20120601TMproductionTM/ We have added 'wirerehang' to the list of case models for plotting in Matlab, so the different cases are now: 1. 'wire' 2. 'fiber' 3. 'erm' 4. 'thincp' 5. 'wirerehang'