2 weeks ago I took some data, and remembered today at the 40m meeting that I hadn't posted it. Bad grad student.
All I'm trying to show here is that we see flashes in the arms that are larger than the ~50 units that we see saturate the Thorlabs transmission PDs. For arm power values below ~50, the QPD sum and Thorlabs PDs give approximately the same values. So, 1 unit on the Thorlabs PDs is equivalent to 1 unit on the QPD sum, and 50 units on the Thorlabs diode is equivalent to 50 units on the QPD sum.
The situation was arms held on resonance with ALS, and the PRMI was flashing.
Arm powers of ~140 imply a power recycling gain of ~7.
Last week, Koji cleaned up the LSC model to make it much more readable, while he was working on piping the ALS signals to the LSC model. However, somehow the DAQ Channels block got deleted before the model was committed to the svn. Since there were 2 months between svn checkins for c1lsc.mdl, it's possible that someone had the model open just to look at, and the block got deleted, and that's the version that Koji started with.
Anyhow, thankfully we have the svn, so Koji and I found that the DAQ Channels block was (as expected) in the previously checked-in version of the LSC model. I put a copy of the old model onto my desktop, opened it up, copied the DAQ Channels block, and then pasted it into the new cleaned-up version of the model. (Jamie - is there a way to conveniently download a previous version through the web interface?)
I have checked it in, compiled and restarted the lsc model. The _DQ channels are back now.
We still have c1mcs having regularly time-over. Can I remove the WFS->OAF connections temporarily?
Yes. I think that's a good idea, since we're not using them at this time, and we want c1mcs to behave. Maybe make an elog note of which version is the first without them, so that we can conveniently find the model(s) in the svn?
I *finally* figured out how to bend Matlab to my will, and close a very simple oplev loop using LQR technology.
This is super, super simple, but it's a step in the right direction. There is no noise, just a simple pendulum with a resonant frequency of 0.75Hz, and a Q of 10. The LQR tries to keep the position of the pendulum at a minimum, and does not care at all about the velocity. You cannot (with Matlab's LQR, at least that I can find) make it care "0" about the control output, so it cares about the control output a factor of 1e-4 as much as the position.
Here are some plots:
The first plot has the open loop system (just the pendulum, no control at all), as well as the closed loop system (the pendulum under control).
Plot 2 is the open loop gain of the controller that the LQR designed.
Plots 3 and 4 are the step and impulse responses of the open loop (pendulum with no control), and closed loop (pendulum with feedback) systems.
The conclusion from the plots (if this were an interesting system) is that it doesn't take much to damp an ideal pendulum. The real conclusion here is that I think I now know how to use the Matlab LQR function, and can make a loop with some noise now.
The ITM oplevs were pretty close to the edge of their ranges, and none of the oplevs have been centered in a while, so I centered ITMX, ITMY, BS, PRM after having done alignment (arms, then PRMI).
Tonight we worked on tweaking up the PRCL new ASC, and then PRMI+1 arm locking. We were unable to get the Xarm to stay locked on a TEM00 mode for very long, and after an hour or two of using the PZTs to try to align the beam to the cavity, we gave up and just used Yarm green.
NB: We haven't done anything to MCL, although it is not in use. Den is still going to get around to elogging what servo shaping he changed on that last night.
I wrote a script that will handle the transitions between the new PRCL ASC and the PRM oplev and local damping. The script is accessible from the PRC ASC screen, and will detect when the PRMI is locked or not. When it is locked, it will turn down the PRM oplev gains and turn on the ASC, and then it will turn off the local shadow sensor damping for PRM pitch and yaw. When the PRMI unlocks, the script will turn off the ASC and restore oplev and local shadow sensor damping.
We saw that the bounce mode of the PRM was getting rung up with our new ASC, so we included a band stop in the ASC, and also turned on the triggering for the PRCL LSC FM6, which has the resonant gain for the bounce mode (as well as roll, and the stack mode). This made the PRMI spot very stable.
We then moved on to green arm locking. The Yarm is behaving perfectly nicely (as nice as it has been lately - it's alignment and mode matching could also use some work), but Xarm was giving us a bit of trouble. As always (since the PZTs were installed?), the mode matching isn't excellent for the green to the arm, so it can be hard to catch a TEM00 mode. Also, even if we did catch a good mode, it would often not stay locked for more than a few tens of seconds. We tried several alignment tweakings, and several different end laser temperatures (within the confines of seeing the beatnote under 100MHz), and didn't have a lot of success. It looks like Eric had the slow servo engaged for the Xend laser, so the temperature offset was something like +300,000, which seemed totally crazy. I turned that off, and found the beatnote somewhere around output of -10,300. So, I haven't gone to the end to look at the temperature, but it's going to be different than when Eric was taking measurements this afternoon. It seems like the main problem with the Xarm is poor mode matching - the maximized input pointing for TEM00, when you unlock and relock the cavity, is just as likely to give you a TEM_9_0 mode, as TEM00.
So, we gave up on the Xarm for the evening, and tried PRMI+1arm, with the new PRCL ASC. This was successful! The Yarm beatnote was around laser slow servo output of +4450. Beatnote at 46.0MHz, -26dBm. We found the IR resonance, moved off, locked the PRMI, transitioned to the new ASC, and brought the Yarm back to IR resonance. What we see is that the power fluctuations in the PRC are much smaller than they were back around Halloween (elog 9338), however the arm power fluctuations still seem very, very large. This is certainly partly due to the fact that we haven't done a thorough Yarm alignment since before messing with the greens, so we will have drifted somewhat. Also, the ALS beatnote sensor isn't perfect, so won't be perfect at holding us near resonance.
Den is thinking about whether we can use the arm transmission QPD signals to feed back to the ETM ASC servos, to try to reduce the RIN in the arms. I feel like we should also see if this amount of power fluctuation can be explained by our ALS noise, because maybe we'll be fine once we transition to IR and turn off the ALS system. Attached is a plot showing that the arm's RIN is coherent with the spot motion seen by the transmission QPD, so we need to check the alignment of the cavity, as well as consider using the trans QPD in an ASC feedback loop.
Here is a plot of the PRC sideband power, as well as the Yarm transmission. The GPS time for this plot is approximately 1070963372.
I noticed that we have not been saving the 1/sqrt(TRX) and 1/sqrt(TRY) data, so I modified the c1lsc model and added them to the DAQ channels block. I restarted the c1lsc model, and the _DQ channels are now archived.
Den just plugged an output from the common mode board into an LSC whitening board (the spare channel that used to be called "PD_XXX_I" in the LSC model). I have modified the LSC model to reflect the new name of the new signal ("CM_SLOW"), and have added this channel to the LSC input matrix. Koji is, I believe, adding this channel to the LSC screen in the auxiliary error signals section. I am also adding the _OUT of the filter bank to the DAQ channels block.
I have looked at the photo of the Xend QPD from elog 9367, as well as the schematic for the board (D990272).
Things that will need swapping out:
I have ordered from digikey via techmart 10 each of the MAX333A's and the OPA140's. (4 per QPD times 2 QPDs plus 2 spares = 10). Both of these new chips have the same footprint and pinout as the part that they are replacing, so it'll be a fairly easy task.
Next up, I need to make a LISO model for the circuit for one of the quadrants, to see what shape it'll turn out to be. Part of this will include deciding what resistors and capacitors to put in the OPA140 gain stage.
Right now, the AD797s say on the schematic that the gain options are different by a factor of 5, but the actual QPD has a different resistor than is on the schematic, and there is also a capacitor in parallel with each resistor, so I need to just pull those out, and pick some values that make sense to me.
Rana and I have discussed ignoring the 2nd and 3rd gain switching options on each quadrant, as that is getting to be more fine control than we need.
Other things on the board:
For now, I will probably just work on the QPD head, and not the whitening board. For now, we can run with 1 stage of whitening, and if we need lower noise, we can revisit the whitening board (including replacing the thick film resistors with thin film).
When thinking about what gains I want on my gain stages, I want to have my full arm power (~700 TRX units) be ~20,000 counts from the ADC. So, I want my single arm power (1 TRX unit) to be ~30 counts from the ADC. This is not such a big number, so this may also require more thinking.
Somehow the POP22 and POP110 demod phases weren't correct anymore. I guess Den saw this after he changed the setup for the REFL165 PD at the LSC rack, but didn't elog it.
I went out to the LSC rack, and found that the power supply that is supplying the amplifiers for both POP22/110 and REFL165 was set to ~16V each channel. I put it back to 15V for each channel. I don't know what Den intended for the 165 amplifier (more volts is more gain), but the POP22/110 amplifier usually runs with 15V.
I also reset the POP22 and POP110 demod phases. Since I'm not able to lock PRMI on sideband this morning (why?!?!), I locked on the carrier, and moved the phases around until POP22 and POP110 were both maximally negative. The phases are/were:
This is a ~60 degree change for both PDs.
I am not sure if Den ever checked the demod phase of REFL165 after he put in the new SMA cable (there's no mention of it in the elog!), so I'm going to check that to see if it helps get PRMI locking back. I know that Den had also been using REFL11 for PRMI locking, but the parameters he used for that aren't in the log either.
I checked out the REFL165 demod phase, and it looks like it was okay. it was -20.9 degrees. I turned on my sensing matrix oscillators, and maximized the PRCL signal in the REFL165_I_ERR channel, and got a pretty good maximization at +155 degrees. I used this to lock the PRMI on sidebands, with MICH gain of +0.3, and PRCL gain of +0.1 .
Since this is working, I'm leaving the REFL 165 phase here, at +155 degrees, although this is almost exactly 180 degrees from what Den left it at, so I'm not sure why I was not able to lock with a demod phase of -20.9. (I tried all 4 permutations of signs, with gain values of the same magnitude (0.3 for MICH, 0.1 for PRCL), and wasn't able to lock. I'll try to figure this out tomorrow, but it was time for the meeting, then the IFO has been busy doing more important things the rest of the afternoon.
Plan for checking: Lock with demod phase of 155, measure TF to one of the other REFL diodes (11, 33 or 55), lock on that other REFL diode. Then, change the REFL 165 demod phase back to -20.9, and measure the transfer function again. Hopefully the answer is just that I was doing something dumb, and it works easily. This test/measurement should only take a few minutes, but it'll make me happier knowing that things still work as they should.
My hunch is that the PRC is SHORT by a few cm, not long.
In my Optickle simulation, the sidebands are not perfectly co-resonating in the PRMI when the arms are not locked. See Fig 1, which is the fields in the PRMI using the design PRC length. If I add 5cm to the PRC length, I get Fig 2, where the peaks are about the same separation, but the upper and lower sidebands have swapped sides of the 0 mark. However, if I remove 5cm from the PRC length, I get Fig 3, where the peaks are much farther apart than in Fig 1. This case looks more similar to the data that Gabriele plotted in his elog entry, where the peaks are separated by at least a linewidth. This is not at all conclusive, but it's a guess for which direction we need to move. Obviously doing an actual measurement will be better.
My tummy feelings also agree with this simulation: When we flipped PR3 (the only optic change in the PRC since Gabriele and I measured the 55MHz peak separation in April), since the HR side of the optic is now at the back, we had to push the whole suspension cage forward to get the beam aligned to the Yarm. Conversely, however, transmitting through the glass substrate adds to the optical path length. So, my tummy feelings may be wrong.
Figure 1, PRC at design length, PRMI sweep:
Figure 2, PRC 5cm longer than design length:
Figure 3, PRC 5cm shorter than design length:
I'm not sure which pointing Rana wanted to fix today, but here's a measurement of the MC spots. They actually look pretty good. There is some room for improvement, but not a lot, so I'm leaving it alone for now, while I play with other things in the IFO.
spot positions in mm (MC1,2,3 pit MC1,2,3 yaw):
[0.63368182839757914, 1.3004245778952557, 0.33621668795755993, -1.5585578137597658, -3.1344594013487286, -1.0533063060089816]
When I got in this morning at 9-something (9:45 maybe?), Steve was taking dust photos on the AS table, of the MC Refl path. Other than that, I don't have any information.
Also, Tuesday is our traditional janitor day, so I'm hesitant to put our blame there. (I think we've kept Tuesdays, even though we're on a less-often schedule....Steve will have to correct me if I'm wrong on this).
The PRCL once again doesn't want to lock on sidebands for me. I can lock on the carrier just fine (using the IFO Config settings, along with some hand-alignment of the PRM).
However, I can't convince it to lock on sidebands. Using the configs that I used on Dec 18th (elog 9491), I'm not getting it. I've done the arm ASS alignment, and I've run LSCoffsets, both of which seemed to do their things appropriately.
I'm going to attribute this today to not being in the groove yet, and I'll look at it again in the morning.
We turned off the WFS servos, and looked at the MC REFL DC, and saw that it was still good, so we said that since the MC spots are pretty good, that we'll keep this alignment for now.
Rana put the beam back on the center of the IOO QPDs on the PSL table.
We switched a steering mirror in the WFS path that was the wrong handed-ness to be the correct handed-ness, then put the beam on the centers of the WFS. We turned on the WFS, and everything seems good.
While we were out on the table, we also changed the anodized aluminum dump for a razor dump, to catch the reflection from the 2inch lens that is the first thing the MC refl path sees out of vac.
There were no major drifts in the WFS error signals while we were gone for dinner, so the MC seems okay for now.
After locking the arms (after the MC alignment work), Manasa and I aligned IP POS, IP ANG, and both end transmission QPDs.
We noticed that IP ANG is clipping in yaw as it comes onto the end table. It looks to me like it's clipping on the edge of the plastic box's aperture, but I can't guarantee that it's not also clipping elsewhere.
The daqd process is segfaulting and restarting itself every 30 seconds or so. It's pretty frustrating.
Just for kicks, I tried an mxstream restart, clearing the testpoints, and restarting the daqd process, but none of things changed anything.
Manasa found an elog from a year ago (elog 7105 and preceding), but I'm not sure that it's a similar / related problem. Jamie, please help us!
Here is a screen dump from the "dtail":
Every 1.0s: dmesg | tail -50 Tue Jan 7 22:43:23 2014
[ 33.498691] [<ffffffff8104a063>] kthread+0x7a/0x82
[ 33.498695] [<ffffffff81003654>] kernel_thread_helper+0x4/0x10
[ 33.498698] [<ffffffff81049fe9>] ? kthread+0x0/0x82
[ 33.498701] [<ffffffff81003650>] ? kernel_thread_helper+0x0/0x10
[ 33.498703] ---[ end trace 6236defa99b3e091 ]---
[ 33.498705] mx INFO: Board 0: allocated MSI IRQ 67
[ 33.498713] mx INFO: CPU0: PAT = 0x7010600070106
[ 33.498715] mx INFO: CPU0: new PAT = 0x1010600070106
[ 33.498718] mx INFO: Board 0: Using PAT index 6
[ 33.499101] eth0: no IPv6 routers present
[ 33.531013] mx INFO: Board 0: device 8, rev 0, 1 ports and 2096896 bytes of SRAM available
[ 33.531017] mx INFO: Board 0: Bridge is 10de:005d
[ 33.531228] mx INFO: Board 0: MAC address = 00:60:dd:46:ea:ec
[ 33.535971] mx INFO: Loaded mcp of len 235448
[ 34.489244] mx INFO: Starting usermode mapper at /opt/mx/sbin/mx_start_mapper
[ 39.148855] mx INFO: mx0: Link0 is UP
[ 39.588511] mx INFO: myri0: Will use skbuf frags (4096 bytes, order=0)
[ 39.589299] mx INFO: 1 Myrinet board found and initialized
[ 287.706367] daqd used greatest stack depth: 3368 bytes left
[86605.907520] daqd: segfault at 38b08e4c0 ip 00007f11b3942a6c sp 00007f10b1917d50 error 4
[86605.907530] daqd: segfault at 38b544f90 ip 00007f11b3942a6c sp 00007f10b12c6d30 error 4 in libc-2.10.1.so[7f11b390e000+14c000] in libc-2.10.1.so[7f11b390e000+14c00
[86605.919454] daqd general protection ip:7f11b3942a6c sp:7f10b1814d30 error:0
[86605.919462] daqd general protection ip:7f11b3942a6c sp:7f10b0bf4d30 error:0
[86605.919615] daqd: segfault at 38aee3db0 ip 00007f11b3942a6c sp 00007f10b0b73d50 error 4 in libc-2.10.1.so[7f11b390e000+14c000]
[86605.919694] daqd: segfault at 38aff35d0 ip 00007f11b3942a6c sp 00007f10b1752d30 error 4
[86605.919701] daqd: segfault at 38b544f70 ip 00007f11b3942a6c sp 00007f10b154dd50 error 4 in libc-2.10.1.so[7f11b390e000+14c000]
[86605.919708] daqd: segfault at 38aff35b0 ip 00007f11b3942a6c sp 00007f10b0ab1d50 error 4
[86605.919733] daqd: segfault at 38b42ae90 ip 00007f11b3942a6c sp 00007f10b10c1d50 error 4 in libc-2.10.1.so[7f11b390e000+14c000]
[86605.919741] daqd: segfault at 38b08e480 ip 00007f11b3942a6c sp 00007f10b0cb6d30 error 4 in libc-2.10.1.so[7f11b390e000+14c000]
[86605.958551] in libc-2.10.1.so[7f11b390e000+14c000] in libc-2.10.1.so[7f11b390e000+14c000]
[86605.958577] in libc-2.10.1.so[7f11b390e000+14c000]
[86605.958586] in libc-2.10.1.so[7f11b390e000+14c000]
[86605.959639] daqd used greatest stack depth: 3160 bytes left
[98139.100888] show_signal_msg: 13 callbacks suppressed
[98139.100895] daqd: segfault at 39c7363b0 ip 00007f5bf253ba6c sp 00007f5b69b48d30 error 4 in libc-2.10.1.so[7f5bf2507000+14c000]
[98687.815120] daqd used greatest stack depth: 2984 bytes left
[208995.594227] daqd general protection ip:7f3b7c930a6c sp:7f3a79f09d50 error:0 in libc-2.10.1.so[7f3b7c8fc000+14c000]
[353015.067479] daqd used greatest stack depth: 2880 bytes left
[367406.863618] daqd: segfault at 41 ip 0000000000000041 sp 00007fb1f0ba2cf8 error 14 in daqd[400000+7c000]
[367406.863833] daqd general protection ip:7fb2f3018a6c sp:7fb1f01c8d30 error:0
[367406.863877] daqd general protection ip:7fb2f3018a6c sp:7fb1f089ad30 error:0
[367406.877408] daqd: segfault at 41 ip 0000000000000041 sp 00007fb1f0ae0ca8 error 14 in daqd[400000+7c000]
[367406.877435] in libc-2.10.1.so[7fb2f2fe4000+14c000]
[367406.877442] daqd: segfault at 39ba287b0 ip 00007fb2f3018a6c sp 00007fb1f034cd30 error 4 in libc-2.10.1.so[7fb2f2fe4000+14c000]
[367406.878372] in libc-2.10.1.so[7fb2f2fe4000+14c000]
[399802.887523] daqd general protection ip:7fb056a71a6c sp:7faf96125f10 error:0 in libc-2.10.1.so[7fb056a3d000+14c000]
[410595.969327] daqd: segfault at 3a91f27b0 ip 00007f48e96eea6c sp 00007f47e6c26d50 error 4 in libc-2.10.1.so[7f48e96ba000+14c000]
[410595.988926] daqd: segfault at 3a91f2790 ip 00007f48e96eea6c sp 00007f47e681bd30 error 4 in libc-2.10.1.so[7f48e96ba000+14c000]
It seems that the most important short-term task we have right now is to figure out what our PRC length is, and what our tolerance from nominal is. Gabriele and EricQ are going to work on that tomorrow. If our PRC is of a length that we can't do anything useful for full IFO locking, we need to open up and fix it sooner rather than later.
While we're in there, we need to also put a baffle on the back side of the PRM cage, to protect the OSEMs from stray light. Den and I discovered before Christmas that turning off the OSEM and OpLev damping to the PRM (while using the POP QPD for ASC) significantly reduced the power fluctuations in the PRC. We still had arm power fluctuations, but I believe those are likely because our ALS system can't hold an arm precisely at full resonance. So, putting a black glass baffle with ~2 inch aperture right up against the OSEMs should help a lot. This week, I'll ask Steve to make me a quickie to-scale cardboard version of the baffles that he has had cut, so I can try securing it to the dirty suspension cage that we have out. I will also check to make sure I have seen with my own eyes the baffles that I need, and copper wire to tie it to the cage.
Other, lower-priority things that we should do eventually:
* Steve, please find another razor beam dump for the WFS reflections - Rana and I used one of the ones that was there for reflection off the 2 inch lens in the MC refl path (replacing the aluminum dump that has been there for ages). We also need to label all of our razor dumps with their purpose, with a label on top, so we remember not to remove dumps that are actually in use.
* At some point, we should change the one remaining steering mirror in the main PSL path that is aluminum, to a steel Polaris ("Polanski" or "Polish") mount. For now, we should just make sure we have one handy. Hopefully this will help reduce the PMC transmission drift that we see.
* Steve, in the morning sometime this week, can you please do a test of the drift of the IOO QPDs? We'd like to see a trend that is maybe 30 or 60 minutes long of the QPD signals. First 10 minutes, all lights in IFO room off. Then, 10 minutes with the lights in the PSL on. Then, the rest of the time the PSL lights off. We want to see if these are hot enough to be causing a big temperature change in the PSL box, which may then be causing some optics to drift.
* QPD code in the simulink models (trans QPDs, but also OpLevs, and anywhere else we do normalization) needs to have anti-divide-by-zero protection. I'll take care of this, it should be a quick copy of what we have elsewhere in the simulink code.
* Note to self for the future, instead of doing a dither alignment for the ASS for the arms, we can use the IP POS and IP ANG, as well as end transmission QPD signals. However, for now, the ASS is working just fine.
* We want to go back to the idea of putting a lens into the in-vac IP ANG path, to avoid the clipping that Manasa and I were seeing tonight. We want something of order 2inch diameter, 1meter focal length. The material doesn't matter, but we do want it AR coated for 1064nm on both sides. We also need to make sure that we could use a fixed 2 inch in-vac mirror mount, or something, to hold this lens. If that won't work, we need to come up with another plan. Manasa is working on thinking about precisely what lens we want to buy for a nice guoy phase telescope for IPANG, so we'll buy a lens after she puts her conclusions in the elog.
* An idea for the MC spots plot that Rana had was to plot the beam tilt and translation, rather than the raw spot positions on the mirrors. The point of this would be to make it easier to see what the output beam from the MC looks like. For MC pointing, we should also think about what our actual tolerances are. The biggest thing is that we need to get through the Faraday without being too close to any edge, and also the REFL beam needs to come back through without clipping. For now, we're just visually checking that the POP beam and the REFL beam both look unclipped since we don't have access to good camera views of either side of the Faraday.
Steve may actually be onto something with the clamps that he had made a year and a half ago. These clamps hold the glass, and then clamp to the base of the suspension cage. Not the table, but the base of the suspension cage. The drawings are in elog 6344. I'm not sure that the 1/4-20 holes in the clamp things are exactly where we'll want them, but we should be able to just dog it down to the base of the suspension. I need to check this, but it may be even easier than tieing the glass to the cage.
Also, something to think about is that the earthquake stop screws extend backwards farther than the OSEMs. I'm not sure anymore if we have shorter 1/4-20 earthquake stops around (if we do, they should be in the cleanroom shelves), but if we can't swap those out, they'll limit how close we can get to the OSEMs.
Here's an overhead photo from 6 Sept 2012:
It looks like there was a 4.4 magnitude earthquake near Fontana, CA around 1:30am today. This tripped all of the suspension watchdogs, which Q has just now re-enabled.
I was in the lab, near the south end of the ITMX oplev table, looking for something, and I bumped the POP ASC QPD's power supply. I thought that it was fine, but did not adequately check it. When EricQ asked me just now about why the PRC is so wobbly today, I checked, and the power for the QPD wasn't properly connected (it's kind of a crappy connector, that if you nudge, contacts or loses contact). Anyhow, I restored power to the QPD, and the PRC looks a little more stable now. My fault for not checking more carefully, and my apologies to Q and Gabriele for their frustrations this afternoon.
Steve noticed that the RGA was not logging data and that not all the correct connection lights were on, and he wasn't able to run the "RGAset.py" script (in ...../scripts/RGA/) that sets up the proper parameters.
I looked, and the computer was not mounting the file system. I did a remote shutdown, then Steve went in and pushed the power button to turn the machine back on. After it booted up, it was able to talk to the file system, so I started ..../scripts/RGA/RGAset.py . The first 2 times I ran the script, it reported errors, but the 3rd time, it reported no communication errors. So, now that the computer can again talk to the file system, it should be able to run the cronjob, which is set to take data at 4am every day. Steve will check in the morning to confirm that the data is there. (The last data that's logged is 22Dec2013, 4am, which is right around when Koji reported and then fixed the file system).
Rana, Gabriele and I are trying to measure the FSR of the PRC (elog about that later), and we turned off the power to the RF generation box so that we could switch cables at the EOM combiner. However, as in elog 9101, the power button won't latch when we try to turn the power back on. All 3 of us tried, to no avail. For our measurement, poor Gabriele is standing holding the button pushed in, so that we can have some RF sidebands.
Tomorrow, we'll have to pull the RF generation box, and put in a better switch.
Since this morning, the fb's timing has been off. Steve pointed it out to me earlier today, but I didn't have a chance to look at it until now.
This was different from the more common problem of the mx stream needing to be restarted - that causes 3 red blocks per core, on all cores on a computer, but it doesn't have to be every computer. This was only one red block per core in the CDS FE status screen, but it was on every core on every computer.
The error message, when you click into the details of a single core, was 0x4000. I elog searched for that, and found elog 6920, which says that this is a timing issue with the frame builder. Since Jamie had already set things on nodus' config correctly, all I did was reconnect the fb to the ntp:
fb$ sudo /etc/init.d/ntp-client restart
As in elog 6920, the daqd stopped, then restarted itself, and cleared the error message. It looks like everything is good again.
I suspect (without proof) that this may have to do with the campus network being down this morning, so the computers couldn't sync up with the outside world.
Steve, please begin the vent!!
We have followed the pre-vent checklist, and done everything except check the jam nuts (which Steve can do in the morning).
We are ready to vent, so Steve, please begin bringing us up to atmosphere first thing in the morning.
Here is a copy of the list, from the wiki:
This sounds great! The only suggestion that I have is for checking POP. If you have the beam on the camera, you can hold a card in front of each mirror, and find out where the edge of the beam is. Introduce the card from the side, and watch for the point where you just start to see the beam on the camera be obstructed. Repeat for the other side, and you have an idea of the centering of the beam.
I think this is most important for the in-vac mirrors, since the beam is large-ish, and we have to hit both steering mirrors at ~45 degrees.
I then used the same settings as in ELOG 9554, except I used -1s instead of +1s for the POP110I trigger matrix elements. (I'm not sure why this is different, but I noticed that the PRC would lock on carrier with positive entries here, so I figured we wanted the peaks with opposite sign).
Nice work!! As with all the other RF PDs, POP110's phase likely needs tuning. You want POP110 (and POP22) I-quadratures to be maximally positive when you're locked on sidebands, and maximally negative when locked on carrier. What you can do to get close is lock PRC on carrier, then rotate the POP phases until you get maximally negative numbers. Then, when locked on sideband, you can tweak the phases a little, if need be.
The PRC (locked on carrier so far today), is pretty wobbly. It'll stay locked on carrier, but it's wobbling. The ASC was over-ridden during the vent. While I was looking around for that, I noticed that the PRM oplev sum is very low.
I went into the lab, and turned off / blocked all oplev beams except the PRM beam. I can't tell what it's clipping on, but there is definitely some red glow in the BS chamber (not as much as the stuff that's coming from the ITMY or SRM oplev hitting a tip tilt suspension - that giant spot went away when I turned off the ITMY/SRM oplev laser). The beam going into the vacuum is nice and strong, but the beam coming out is very weak, and has a horizontal line of scatter through it, like it's clipped somewhere in pitch. The PRM oplev sum is currently ~150 cts, when it should be closer to 2,000.
So far, this seems to be livable, but it's definitely disappointing.
In an effort to stop the PRC from wiggling around so much, I recentered the POP QPD after maximizing the POPDC power when locked on carrier. The beam was basically off the QPD in yaw, and at half-range in pitch.
I have turned off the 3.2Hz res gains in the PRC ASC loops, since those seem to make the loops unstable.
Right now the pitch gain is -0.001, with FM1,3,9 on. Yaw gain is -0.004, with FM1,3,9 on.
Pitch gain can't increase by factor of 2 without oscillating.
I tried to take transfer functions, but I think the ASC situation is really confusing, since I have OSEM damping, oplev damping, and this POP QPD damping on the PRM. It's hard to get coherence without knocking the PRC out of lock, and it keeps looking like my gain is 0dB, with a phase of 0 degrees, from ~1 Hz to ~10 Hz. Outside that range I haven't gotten any coherence. Moral of the story is, I'm kind of puzzled.
Anyhow, as it is right now, the ASC helps a bit, but not a whole lot. I increased the trigger ON value, so that it shouldn't kick the PRM so much. I wish that I had implemented a delay in the trigger, but I'm not in the mood to mess with the simulink diagrams right now.
[Rana, Jenne, Manasa]
We looked at the I vs. Q separation in several of the Refl PDs, while driving the PRM, while the PRMI was locked on sidebands.
For REFL 55, we adjusted the demod phase to try to minimize the peak in the Q signal, and were only able to get it to be about 1/10th the size of the I peak. This is not good, since it should be more like 1/100, at least.
For both REFL 11 and REFL 165, we were able to get the Q peaks to less than 1/100 of the I peak height.
We changed the REFL55 phase from 17 to 16, and the REFL165 phase from -160.5 to -162.5.
Since we believed that we had done a good job of setting the demod phase for REFL165, we used it to also check the balance of BS/PRM for MICH locking. I drove the BS with an arbitrary number (0.5), which creates a peak in the I phase of REFL165, and then I put in a drive on the PRM and tweaked it around until that peak was minimized. I came up with the same ratio as Koji had last Friday: BS=0.5, PRM=-0.2625. (The old ratio we were using, up until ~December when we started locking MICH with the ITMs, was BS=0.5, PRM=-0.267).
Also, while we were locked using REFL55 I&Q, we noticed that the other REFL PDs had lots of broadband noise in their I signals, as if some noise in the REFL55 diode is being injected into the PRM, that we are then seeing in the other PDs.
Some checks that we need to do:
* Inject a calibration line, set all the peak heights equal, and look at the noise floors of each PD.
* Use the calibration line to calibrate the PDs (especially REFL165) into meters, so that we know that it's noise is low enough to hold the PRC through the CARM offset reduction.
* Check out the state of the transmission QPDs - what is their noise, and is it good enough to use for holding the arms after we transition from green beatnote locking? Does the whitening switching do anything? What is the state of the whitening?
I'm not happy with the beam position on that first lens, but since it's so crazy in the BS chamber, and the PRM oplev has something like 5 in-vac steering mirrors, I'm hesitant to suggest that we do anything about it until our next vent. But we should definitely fix it.
I calibrated the REFL signals to meters from counts. The I-phase signals all line up very nicely, but the Q-phase signals do not at all. I'm not sure what the deal is.
I locked the PRMI on sidebands, and drove the PRM. I looked at the peak values at the drive frequency in the REFL signals, and used that as my "COUNTS" value for each PD.
I know the PRM actuator calibration is 19.6e-9 (Hz/f)^2 m/ct , so if I plug in my drive frequency (564 Hz, with the notch in the PRC loop enabled), and multiply by my drive amplitude in counts, I know how many meters I am pushing the PRM. Then, to get a meters per count calibration, I divide this calibration number (common for every PD) by the peak value in each PD, to get each signal's calibration.
As a side note, I also drove MICH, and tried to use this technique for the Q-phase calibrations, but neither calibration (using the PRCL drive nor the MICH drive) made the Q-phase signals line up at all.
At least for the I-phase signals, it's clear that REFL33 has more noise than REFL11 or REFL165, and that REFL55 has even more noise than REFL33.
Here are the calibration values that I used:
I was able to get the PRMI locked on REFL33 I&Q, but it wasn't overly stable, since there is so little separation between the MICH and PRCL signals in that PD.
We have already adjusted the phase to maximize PRCL in the I-phase. Since MICH is ~45 degrees separated from PRCL, there is some projection of MICH in the I-phase, and some in the Q-phase.
To remove this MICH component, I locked the PRMI on REFL55, and drove MICH. I looked at REFL33I at the CARM filter bank input (as just a dummy location to get a signal into DTT). I then added REFL33Q to the CARM row of the input matrix, to try to get the MICH line minimized. I then used these values for PRCL, and used just REFL33Q for MICH, and re-locked the PRMI. The PRMI was much more stable and happy.
The input matrix values that I used were:
MICH: REFL33Q = -0.487, Servo Gain = -20.0
PRCL: REFL33I = 1.556, REFL33Q = 1.8, Servo Gain = -0.020
Some locking notes:
The PRMI is very sensitive to alignment, and the PRM tends to drift away from optimal alignment on a ~1 hour timescale. When the PRM was not well aligned, it looked like MICH had a locking offset (manifested as non-equally sized blobs at AS). The MICH offset seemed to go away when we realigned the PRM.
[Koji, Jenne, EricQ, Manasa]
We had a short discussion this evening about what our game plan should be for transitioning from using the ALS system to IR-generated error signals.
The most fundamental piece is that we want to, instead of having a completely separate ALS locking system, integrate the ALS into the LSC. Some time ago, Koji did most of the structural changes to the LSC model (elog 9430), and exposed those changes on the LSC screen (elog 9449). Tonight, I have thrown together a new ALS screen, which should eventually replace our current ALS screen. My goal is to retain all the functionality of the old screen, but instead use the LSC-version of the error signals, so that it's smoother for our transition to IR. Here is a screenshot of my new screen:
You will notice that there are several white blocks in the center of the screen. From our discussion this evening, it sounds like we may want to add 4 more locking servo paths to the LSC (ALS for each individual arm, and then ALS for CARM and DARM signals). The reason these should be separate is that the ALS and the "regular" PDH signals have different noise characteristics, so we will want different servo shapes. I am proposing to add these 4 new servo blocks to the c1lsc model. If I don't hear an objection, I'll do this on Monday during the day, unless someone else beats me to it. The names for these filter modules should be C1:LSC-ALS_XARM, C1:LSC-ALS_YARM, C1:LSC-ALS_DARM and C1:LSC_ALS_CARM. This will add new rows to the input matrix, and new columns to the output matrix, so the LSC screen will need to be modified to reflect all of these changes. The new ALS screen should automatically work, although the icons for the input and output matrices will need to be updated.
The other major difference between this new paradigm and the old, is the place of the offset in the path. Formerly, we had auxiliary filter banks, and the summation was done by entering multiple values in the ALS input matrix. Now, since there is a filter bank in the c1lsc model for each of the ALS signals precisely where we want to add our offsets, and I don't expect us to need to put any filters into those filter modules, I have used the offset and TRAMP of those filter banks for the offsets. Also, you can access the offset value, and the ramp time, as well as the "clear history" button for the phase tracker, all from the main screen, which should help reduce the number of different screens we need to have open at once when locking with ALS. Anyhow, the actual point where the offset is added has not changed, just the way it happens has.
When we make the move to using the ALS in the LSC, we'll also need to make sure our "watch arm" and "scan arm" scripts are updated appropriately.
As an intermediary locking step, we want to try to use the ALS system to actuate in a CARM and DARM way, not XARM and YARM. We will transition from using each ALS signal to feed back to its own ETM, to having DARM feed back to the ETMs, and CARM feed back to MC2. We may want to break this into smaller steps, first lock the arms to the beatnotes, then find the IR resonance points. Transition to CARM and DARM feedback, but only using the ETMs. After we've done that, then we can switch to actuating on MC2. If we do this, then we'll be using the MC to reduce the CARM offset.
Once we can do this, and are able to reduce the CARM offset, we want to switch CARM over to a combination of the 1/sqrt(transmission) signals. The CARM loop has a tighter noise requirement, so we can do this, but leave DARM locked to the beatnotes for a while.
After continuing to reduce the CARM offset, we will switch CARM over to one of the RF PDs, for its final low-noise state.
We'll then do a quick swap of the DARM error signal to the AS port (maybe around the same time as CARM goes over to a PDH signal, before the CARM offset is zero?).
During all of this, we hope that the vertex has stayed locked. If our 3f sensing matrix elements are totally degenerate when the arms are out of resonance, then we may need to acquire lock using REFL 1f signals, and as we approach the delicate point in the CARM offset reduction, move to 3f signals, and then move back to 1f signals after the arm reflection has done its phase flip. Either way, we'll have to move from 3f to 1f for the final state.
Facilities just came by and cleaned the smoke detector that is above Steve's desk. It's next to an air vent, so I guess it collects dust more than a "typical" smoke detector.
As we've been seeing a bit lately, the MC will be locked happily for several hours, but then it will start misbehaving.
Today, I measured the spots on the MC mirrors, and found that the MC2 spot was quite far off in yaw (about -3.5 cm). I recentered the MC2 spot, and then (with the MCWFS on), moved MC1 and 3 until their WFS outputs were close to zero (they had gone up to 100+). In the ~15 minutes since doing that, the MC refl signal is not oscillating like it was, the transmission is up, and the MC has not unlocked.
To reiterate, I did not touch any settings of anything, except the alignment of the MC mirrors to center the MC2 spot, and then offload the WFS. Next time the MC starts acting up, we should measure the spots, and roughly center them, before messing with any other settings. Note however, that this is a ~10 minute procedure (including the fact that one spot measurement takes a little less than 5 minutes). This need not be a several hour endeavour.
Koji mentioned to me (and elogged) that he was unsuccessful locking the ALS using the LSC servos. He suggested I look into this.
So, rather than just looking at the transfer function between POX or POY and the green beatnotes at a single frequency, I did a whole transfer function. The point was to see if the TF is flat, and if we get any significant phase lag in the transfer from c1als to c1lsc. (c1als is running on the IOO machine, so an RFM connection is involved in getting it over to the LSC machine.)
In the first figure, I have plotted POX vs. Beatnote_PHASE_OUT (ALS error signal, still in the c1als model), and POX vs. ALSX_IN1 (the ALS error signal, after transfer over to the c1lsc model). You can see that we have a little phase lead in the blue transfer function, and fairly significant phase lag in the red (red is after transfer over to the lsc model). In the grand scheme of things, the magnitude is fairly flat, however that is not perfectly true - the peaks seen near 50 Hz and 300Hz are repeatable. The relative phase lag between the "BEATX" version of the signal in the ALS model, and the "ALSX" version of the signal in the LSC model is 15 degrees at 200 Hz, which corresponds to 33 usec.
The second figure is the same as the first, except for the Yarm. The relative phase lag between the ALS version of the error signal and the LSC version is 16 degrees at 200 Hz, which is about 35 usec.
As a side note, before trying any ALS locking, I took a spectrum of the beatnote (in the ALS model) while the arms were locked with IR:
To check things, I made sure that I could lock the Xarm ALS using the old ALS system - I was able to do so. (Has someone put the "watch" script as a constantly-on thing? It's kind of nice not to have to turn it on, although we'll need to change it to turn off the LSC versions of the servos eventually).
Then, I tried locking the Xarm using the LSC system (using only FM5 of the regular LSC-XARM filter bank). Like Koji, I was not able to acquire lock. As a next step, I copied all of the LSC-XARM filters into an empty filter module, LSC-XXXDC (the first one on the list underneath LSC-XARM), and copied over the ALS Xarm filters to the LSC Xarm filter bank. I then tried to acquire lock, but am unable to get it to stay. Using the ALS system, when you put in a small gain, the beatnote starts to settle down, and as you increase the gain, the beatnote stops moving (as seen on the spectrum analyzer) almost completely. However, using the LSC system, the beatnote never really stops moving or settles down. And if I increase the gain, I push the ETM hard enough that I lose green lock. I have put the regular LSC filters back for now.
Here is a plot from Foton comparing the FM5 filter modules from the LSC-XARM (regular IR locking) and the ALS-XARM servo. They are pretty different, and have 10 degrees of phase difference at 200 Hz, because 2 of the 3 poles are complex in the LSC version, while the ALS version is just a single real pole.
Anyhow, I am declaring it to be dinnertime, and I plan to return in a few hours. Since I put the regular LSC filters back (since I'm going to have to realign after dinner anyway), the IFO should be in its nominal state if anyone wants to come in and play with it.
It looks like its somehow a discrepancy between the TFs of each error signal, because features are similar, and present, in both error signals.
I'm really excited, so I'm posting this, even though I'm still working:
I currently have ALS locked using the LSC system, and have (by hand, coarsely) found IR resonance! Hooray!
I looked at my error signals, as well as LSC-XARM_IN1 with dataviewer, and noticed that the XARM_IN1 signal was crazy when I was using the ALS signal as the error. I soon realized that this is because there was a non-zero element in the power normalization matrix, and I'm overriding the trigger. So, I was trying to divide by zero, and was getting crazy numbers. After zeroing the power normalization matrix element for the Xarm, the XARM_IN1 signal matched the ALSX_OUT, and I was easily able to acquire lock.
I had already re-transferred over the ALS versions of the filters, so that's what I'm using right now. Next up (on a 5 minute time-scale) is trying to acquire lock using the regular LSC filters.
Oh, also, something I hadn't thought of before dinner: I am setting the offset of the ALSX filter bank such that the output is centered around zero, so that I can lock, since these are not AC coupled servos.
I am also not able to lock the ALS using the 'regular' LSC filters. To figure out what filters were doing what, I made several comparison plots from Foton.
The first one is the progression of ALS locking, using the filters from ALS-XARM. FM5 is always engaged, then FMs 2, 3, 6, 7, and 8, and finally FM 10 (the low frequency boost) is engaged.
The next plot is a comparison between the ALS version of the filters, and the LSC-XARM equivalents.
Finally, just so I remember which LSC filters do what, I made an equivalent of the first plot, but for the LSC filters.
When I try to lock the Xarm ALS using the regular LSC filters, I'm getting an oscillation somewhere, that grows and eventually knocks me out of lock. It looks from dataviewer to be in the ~few Hz range, but it's hard to see it in DTT, since I don't stay locked all that long once the oscillation starts. (If I catch it, I can back off the gain and turn off the servo without losing lock, but if I don't turn off the servo, I inevitably push the ETM too hard and lose green lock to the arm.) I tried engaging the 3.2 Hz resonant gain filter, and it just makes things oscillate sooner, so that's not a solution with the current filter designs.
Also, I'm not able to lock the IR using the ALS version of the XARM filters. I'll have to meditate more on the situation, but the filters seem to be different enough that there's no crossover at this point.
No more progress tonight. I am still unable to lock the ALS using the regular LSC filters. I went back to putting the ALS filters into the LSC filter banks, and locked both arms with ALS, and found their IR resonances. I then held them off resonance, and tried to lock PRMI with REFL 55 I&Q, with no success. Just before locking the arms, I had redone the whole IFO alignment (lock arms in IR, ASS, lock and align MICH, lock and align PRMI), and the PRMI was flashing very nicely. I'm not sure why I wasn't able to catch lock, except that perhaps 3 or 6 ALS offset counts isn't far enough away from the IR resonance to make the 1f signals happy. The MC lost lock, which I then took as a sign that it's time to go home. (I was hoping to do a quick PRMI + 2arms, and see that we don't lose PRMI lock. I was going to catch lock with REFL55, then transition to REFL33, although if I had thought about it before the MC lost lock, I would have tried just catching lock with REFL33).
I restored the regular LSC filters for the X and Y arms, and locked the arms in IR just to make sure it's all honkey-dory. Which, it's not quite. I don't know why, but right now, neither arm wants its boost (FM9) enabled. It's part of the restore script that FM9 is triggered along with the rest of the filters, but even if I turn on the filters manually, I can turn on all but FM9, and then when I turn on the boost, the arm falls out of lock. Same behavior for both arms. Anyhow, they lock, and they seem okay modulo the boost not being able to engage.
We need to change several scripts for use with the new ALS-in-the-LSC paradigm:
* Watch arms (to turn off ALS if we lose the beatnote, before pushing optics too hard)
* Find IR resonance
* Offset from resonance
None of these should be difficult, just changing the filter bank names to match the new ones (ex. LSC-XARM rather than ALS-XARM, and LSC-ALSX rather than ALS-OFFSETTER1).
So far, I have changed the "find resonance" script (ALSfindIRresonance.py). I believe, in principle, to first order, that my modifications should work, however I have not yet tested the script. So. If you use it, watch the output of the script and ensure it's doing what it ought. I'll check it after the lunch meeting and update this log entry. (I changed the name of the "OFSFILT" variable, line 26, and also modified line 114. Both of those lines have comments on how to revert the changes).
I have also changed the "offset from resonance" script (ALSchangeOffset.py). Again, since I'm not locking right now, I have not tested this script either. So, pay attention if you need to use it, before I check it. (I changed the name of the OFSFILT variable, and the check which arm logic around line 37. Again, both of those lines have comments on how to revert the changes.)
[Jenne, Koji, Manasa, EricQ]
Today we successfully locked the ALS using the LSC system, with filters that are good for both the IR PDH and the ALS locking. We tried PRFPMI, but were unable to hold PRMI lock while the arms were held with ALS. We combined the ALS signals into common and differential signals, and successfully transitioned over to a combined set of 1/sqrt(TRANS) signals for the common mode part of the lock (differential stayed with ALS).
Locking the ALS using filters in the LSC system that are also good for IR PDH
The biggest difference between the ALS and LSC filters were the ones used for lock aquisition. At Koji's suggestion, I made FM5 of the LSC servos (for X and Y arms) the filter needed for ALS locking. Then, I made FM4 into a combination of old LSC FM4 and FM5, as well as an inverse of the new FM5, so that when both FM4 and FM5 are engaged, the servo shape is the same as the old LSC. I left the other LSC filters where they were. I replaced the FM1 +6dB with the combined integrators (really, just gentle DC boosts) for the ALS, since we were never using this +6dB filter module. The LSC resonant gain filter for the bounce mode also included a resgain for 18.5 Hz. I don't know what that was for, and it was eating into phase that I needed, so I removed it.
The other filter that changed significantly was the Boost filter. The ALS system had been using more DC gain than the LSC had. However, the current ALS boost filter (in FM10 of the old ALS servos) was eating too much phase near my UGF. So, I scooted the whole boost filter to lower frequencies, to give myself some extra phase margin. The boost was set to "zero history", "zero crossing", with 0.01 tolerance and an 8 second timeout. Setting it to zero crossing with a low tolerance, rather than just ramping it on, was the key to engaging the boost.
I had to be so careful about phase margin, since I lost ~15 degrees of phase at 200 Hz from the lag of going through the RFM network. This was pretty frustrating, but I don't have a better plan yet, save moving the c1als model and ADC to the SUS machine, which has Dolphin access to the LSC. I may back off my safety margin, and give myself some gain in the boost back at 10Hz, since we are now seeing too much noise at 10Hz in the closed-loop spectra. I also "cheated" and lowered my UGF from the ~150Hz it used to be in the ALS model, to 100Hz, where I was closer to the top of the new phase bubble.
With the new filter situation, I was able to lock the Xarm (the one I was using for design work) with both IR and ALS. To lock IR, the "restore" script still works. For the ALS, we should put in a separate "restore" script into the IFO_CONFIGURE screen.
The ALS locking procedure is as follows:
* Prepare ALS and green locking. Green locked to 00 mode, alignment all nice, etc, etc. Beatnote within 100MHz on spectrum analyzer. If doing both arms, try to get beatnotes on opposite sides of PSL, to keep crossbeatnotes at higher frequencies, and out of the way.
* Turn on Watch script.
* Set LSC parameters (this is where a new restore script will come in handy):
* Zeros in RFPD columns of input matrix (i.e. POX and POY).
* Ones in AUX input matrix elements.
* Zeros in power normalization matrix rows for arms.
* All FM triggers for arms set to "Man" for manual.
* Override main trigger, so that signals are always going through to the servo.
* Only FM5 engaged in arm servo.
* Gain of servo set to zero, output on, then engage main LSC master switch. ETM output on.
* Clear history in phase tracker.
* Check sign of gain using + or - 0.1 in the servo. You'll know if you got it wrong (the ETM will be kicked, and the beatnote will fly around). If you didn't get it wrong, you probably got it right.
* Increase gain to about 12 (with correct sign).
* Engage FM1 (gentle DC boost), FM6,7,8 (resonant gains for stack, bounce, roll)
* Wait a few seconds for filters to settle, then engage FM9 (boost).
* Run find IR resonance script.
* Move off resonance by ~36 counts (12 times the +3 script). This number comes from trying to be completely off the IR resonance, even when the PRMI was locked.
* Do whatever locking (ex. PRMI) you set out to do.
After locking both arms with ALS using the LSC system, we attempted to lock the PRMI. We were able to lock PRMI on REFL55 I&Q, REFL33 I&Q, and REFL55 I&AS55Q before the arms were locked, so we were hoping that we wouldn't have too much trouble.
We found the IR resonance for both arms, then moved off resonance. Then, restored the PRM. For REFL55, Koji coarsely turned the REFL 55 demod phase from 16 degrees to 87, while we were locked on the carrier. After this, I stepped farther and farther from the IR resonance, since at first I found that our transmitted powers were something like 4, rather than almost zero, so the demod phase may not be totally correct.
We were having trouble, so we locked the PRMI on carrier using REFL55 I and AS55 Q, with 1's in both elements in the input matrix. MICH gain was about -10, PRCL +0.010. We used this time to tweak up the alignment of the PRMI. At some point, Koji tweaked the REFL33 demod phase from 124 to 134 degrees. Then we switched back to sideband locking. After some trials with REFL55 I&Q, and REFL55/AS55, we went to REFL33 I&Q. REFL33I->PRCL was 1.556 in the input matrix, and REFL33Q->MICH was -0.487. No other elements in the input matrix. MICH gain was reduced to -6, PRCL gain to -0.020. MICH FMs 3,6,9 triggered, PRCL FMs 2,3,6,8,9 triggered. We were able to keep short locks on the order of ~10 seconds, but not longer. We played with every parameter we could think of (alignment being good is one of the most important!), but were not able to keep better lock. The POP spot is moving around a lot, so the PRCL ASC needs to be examined, hopefully tomorrow.
We started losing the Xarm lock fairly regularly, I'm not sure why, but the Yarm was locked for almost 2 hours straight, held off resonance with ALS!
ALS Common and Differential, transition to IR control
We set PRMI aside for the rest of the night, and looked at using ALS to control the arms in common and differential modes.
Regular ALS locking procedures were used (see above), with the exception of the AUX input matrix:
Since the beatnotes were on opposite sides of the PSL frequency, the common and differential modes look opposite of what you'd expect.
We then used the regular find IR resonance scripts running simultaneously, which worked really well to find both arms' IR resonance points.
I put a 1 count offset in the Xarm servo (which was our proxy for common mode), although in retrospect this should have been +0.5 in ALSX, and -0.5 in ALSY, so that our signals going through the input matrix were at their zero crossings. Anyhow, this offset put us at about half fringe on both arms (transmissions were about 0.6).
Koji set the offsets in the 1/sqrt(trans) filter banks before the input matrix so that they would have zero crossings at this point (avg the IN1, put negative of that value into the offset).
We then stepped the input matrix values until our common mode (Xarm) row was:
We left the differential (YARM) row alone, so that the ALS system would still be controlling the differential degree of freedom. The values and sign for the 1/sqrt(trans) signals came from a transfer function of dividing the spectra of each error signal and noting the relative gain and sign.
After we swapped the error signals, we realized that we had to remove the offset from the XARM servo, which is why we should have put the offsets elsewhere in the first place.
Then, Koji took a spectrum, which is attached to this entry. We note that the ALS signals are strongly correlated, and mostly common.
To Do List
Going forward, we need to figure out what is going on with the PRMI, and why we're having trouble keeping lock.
We need to redo the PRCL ASC servo, with the anti-oplev trick that Rana mentioned a week or two ago.
We need to investigate the degeneracy of REFL165, now that Q's simulation doesn't justify / explain it.
I measured the REFL 165 demod board's I/Q separation.
Our 11MHz signal is currently 11.066092 MHz, so I put a signal to the RF input of the REFL165 demod board at 165.992380 MHz (15*11 MHz + 1kHz), with a signal of -13 dBm.
I then used the SR785 to measure the transfer function between the I and Q output channels.
I got 82.7 degrees, at -0.64 dB. (I don't remember now if I had I/Q, or Q/I, not that it really matters). So, it seems that the REFL165 demod board has good separation, and at least isn't totally broken.
I noticed that the fb lights on all of the models on the c1lsc machine are red, and that even though the MC was locked, there was no light flashing in the IFO. Also, all of the EPICS values on the LSC screen were frozen.
I tried restarting the ntp server on the frame builder, as in elog 9567, but that didn't fix things. (I realized later that the symptom there was a red light on every machine, while I'm just seeing problems with c1lsc.
I did an mxstream restart, as a harmless thing that had some small hope of helping (it didn't).
I logged on to c1lsc, and restarted all of the models (rtcds restart all), which stops all of the models (IOP last), and then restarts them (IOP first). This did not change the status of the lights on the status screen, but it did change the positioning of some optics (I suspect the tip tilts) significantly, and I was again seeing flashes in the arms. The LSC master enable switch was off, so I don't think that it was trying to send any signals out to the suspensions. The ASS model, which sends signals out to the input pointing tip tilts runs on c1lsc, and it was about when the ass model was restarted that the beam came back. Also, there are no jumps in any of the SOS OSEM sensors in the last few hours, except me misaligning and restoring the optics. I we don't have sensors on the tip tilts, so I can't show a jump in their positioning, but I suspect them.
I called Jamie, and he suggested restarting the machine, which I did. (Once again, the beam went somewhere, and I saw it scattering big-time off of something in the BS chamber, as viewed on the PRM-face camera). This made the oaf and cal models run (I think they were running before I did the restart all, but they didn't come back after that. Now, they're running again). Anyhow, that did not fix the problem. For kicks, I re-ran mxstream restart, and diag reset, to no avail. I also tried running the sudo /etc/init.d/ntp-client restart command on just the lsc machine, but it doesn't know the command 'ntp-client'.
sudo /etc/init.d/ntp-client restart
Jamie suggested looking at the timing card in the chassis, to ensure all of the link lights are on, etc. I will do this next.
The LSC machine isn't any better, and now c1sus is showing the same symptoms. Lame.
The link lights on the c1lsc I/O chassis and on the fiber timing system are the same as all other systems. On the timing card in the chassis, the light above the fibers was solid-on, and the light below blinks at 1pps.
Koji and I power-cycled both the lsc I/O chassis, and the computer, including removing the power cables (after softly shutting down) so there was seriously no power. Upon plugging back in and turning everything on, no change to the timing status. It was after this reboot that the c1sus machine also started exhibiting symptoms.
Koji noticed that the time on the front-end detail screens was not correct, and that the GPS time was not matching up between different models. Koji ran the following on all front-end machines, and on nodus:
sudo ntpdate -b -s -u pool.ntp.org
Now, everything is fine, and every status light on the cds overview screen is green.
I have recalibrated the REFL signals.
I first adjusted the demod phases until the I-signals lined up with the I-phase in the sensing matrix plot:
I then balanced the ITM drives by pushing on -1*ITMX and +1.015*ITMY, and seeing a minimum of MICH actuation in the I-phase of REFL55 (the PD I was locking with).
I then took a nice long measurement with DTT, and measured the peak heights in I and Q for each REFL diode. I was driving PRM with 100 cts at 675.1Hz, and ITMX with 1000 cts at 452.1 Hz (and matching ITMY drive, to make pure MICH). Knowing these numbers, and the actuator calibrations (PRM elog 8255, ITMs elog 8242), I know that I was driving PRCL by ~4.3 pm, and MICH by ~23 pm.
For the I-phase calibrations, I find the peak height at the PRCL drive frequency, and divide 4.3 pm by that height. For the Q-phase calibrations, I find the peak height at the MICH drive frequency, and divide 23 pm by that height.
This gives me the following calibrations:
Calibration [picometers / count]
REFL 11 I
REFL 11 Q
REFL 33 I
REFL 33 Q
REFL 55 I
REFL 55 Q
REFL 165 I
REFL 165 Q
My calibrated REFL spectra then looks like: