Detector Watch S6 V1

ML1
ML1
Joined: 20 Feb 05
Posts: 347
Credit: 86307557
RAC: 604

RE: Well, at least they're

Message 93804 in response to message 93802

Quote:
Well, at least they're not letting the weather get them down at Hanford :
Quote:

Sci Mon Summary (no science mode this shift)

[...]

22:00 UTC Had a keek ootside and it's still minging.
(I had a look outside, and it's still not very nice)

23:00 UTC Totally Scunnered Man!
(I am a little fed up)

Here endeth today's Scottish lesson...


You sure they ain't been infiltrated by a Feegle or few from Terry Pratchett's Wintersmith?...

The wind pressure induced movements of the buildings and the ground should be a good test of how well the active supports operate. Or are they suffering from the effects of air pressure changes also?

You could even make a good fun sci-fi horror movie out of a night like that! Revenge of the tumble weed? Lightning bolts down the tubes? St Elmo's fire across the arms?

Cheers,
Martin

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

ML1
ML1
Joined: 20 Feb 05
Posts: 347
Credit: 86307557
RAC: 604

RE: I'm still reading to

Message 93805 in response to message 93802

Quote:
I'm still reading to define what it is that is upsetting the locks. So far it seems the new setup - higher input power, output mode cleaner and TCS - make the control systems more twitchy.


Increased effective gain on the detectors due to the increased laser power that is then pushing the feedback loops into instability?... (Sorry if that aspect is too 'obvious' and you're looking at what aspects of the feedbacks is pushing into instability.)

Regards,
Martin

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

RE: You sure they ain't

Message 93806 in response to message 93804

Quote:
You sure they ain't been infiltrated by a Feegle or few from Terry Pratchett's Wintersmith?...


Crivens!! Oh, waily, waily, waily!! Is that an egg, or a major root vegetable question? :-) :-)

[ They just need to leave some of Granny Aching's Special Sheep Liniment out at night. I feel all will settle down if so. ]

Actually you are closer than you think, this was posted on the Hanford ilog :

Quote:
The wind pressure induced movements of the buildings and the ground should be a good test of how well the active supports operate. Or are they suffering from the effects of air pressure changes also?


I go on the principle : "if there's energy, it might get in".

Quote:
You could even make a good fun sci-fi horror movie out of a night like that! Revenge of the tumble weed? Lightning bolts down the tubes?


Do it in black and white, with a guy in a rubber suit, and then blame it on global warming. I guess the LHC is the last time any experimenter is going to say 'black hole' in public. :-)

Quote:
Increased effective gain on the detectors due to the increased laser power that is then pushing the feedback loops into instability?... (Sorry if that aspect is too 'obvious' and you're looking at what aspects of the feedbacks is pushing into instability.)


Yup, I was after the detail. They have this thing called a Pockel's cell at the input - between the laser and the power recycling cavity - that modulates the intensity at radio frequency. That is picked up at various points in the interferometer and becomes the input to the feedback circuits that control the mirror positions. The output mode cleaner ( OMC ) is there to remove that RF, and other 'stray' stuff, so that the 'science' photodiode gets a more pure differential arm signal. The OMC is a new feature, in previous runs the RF component was part of the science numbers. Both interferometers seem to have been talking alot about the OMC, with the recent difficulties.

At Livingston there are other theories. Here is a 'glitch' :

which

Quote:
..I would have to guess that it was caused by some sort of pidgeon or robot lobster :P


Cheers, Mike.

( edit ) I'm not entirely sure if this is tongue in cheek :

Quote:
20:47:34 UTC Lost Lock
End of segment 513
Reason: Nothing clear on seismic in the 0.2 or 2 Hz scans, a very slight increase in EY at the time, so it could be the wind I mentioned just before it lost lock. Changed the frequency range in the script and there is seismic showing up at EY in 10-50 Hz.
(I did, however, also spot two bunnies hopping on the grass outside the kitchen shortly after this, so it could have been their hopping).


that is : what do bunnies do at 10 - 50 Hz ? :-)

[ or 0.2 - 2 Hz? Needs more data ..... ]

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

ML1
ML1
Joined: 20 Feb 05
Posts: 347
Credit: 86307557
RAC: 604

RE: that is : what do

Message 93807 in response to message 93806

Quote:
that is : what do bunnies do at 10 - 50 Hz ? :-)


Quiver in anticipation or do they really do it that fast?!

Quote:
[ or 0.2 - 2 Hz? Needs more data ..... ]


For their size and mass, that is far more plausible.

Or is the wistful stare out the window just a sign of frustration with the ultra-sensitive equipment and all the unwanted external interference?

You certainly have difficulties when even an air suspended optics bench is itself too noisy!

Crivens!! Indeed!

Perhaps something else far too obvious and 'silly'... Perhaps try reducing all the detector dependant feedbacks gain in proportion to the laser intensity?... (Or is that supposed to have been done already?)

Or is the problem elsewhere and dependent directly on the small effects of heating from the laser beam? (Or anything of anything else!?)

I'm sure you could make a good thriller just out of the LIGO tune-ups... To then finally and slowly tune into... something... astronomically unexpected...

Add in some occult secret society trying all manner of obscure political and physical and occult tricks to try to secretly sabotage the tune up all to keep the secrecy of their great secret... And... Adopt your own Alpine Kat... And...

We should have something far more interesting than anything Dan Brown can plagiarise!

;-)

Regards,
Martin

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

ML1
ML1
Joined: 20 Feb 05
Posts: 347
Credit: 86307557
RAC: 604

RE: ( edit ) I'm not

Message 93808 in response to message 93806

Quote:

( edit ) I'm not entirely sure if this is tongue in cheek :

Quote:
20:47:34 UTC Lost Lock
End of segment 513
Reason: Nothing clear on seismic in the 0.2 or 2 Hz scans, a very slight increase in EY at the time, so it could be the wind I mentioned just before it lost lock. Changed the frequency range in the script and there is seismic showing up at EY in 10-50 Hz.
(I did, however, also spot two bunnies hopping on the grass outside the kitchen shortly after this, so it could have been their hopping).


It certainly says something for the sensitivity if you can't be sure about the bunny hops!

Or is there more of a concern about where they are tunneling?...

Regards,
Martin

See new freedom: Mageia Linux
Take a look for yourself: Linux Format
The Future is what We all make IT (GPLv3)

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

RE: It certainly says

Message 93809 in response to message 93808

Quote:
It certainly says something for the sensitivity if you can't be sure about the bunny hops!


This is a good all round description of the LIGO design thinking circa 1996. I was a bit surprised to read this ( page 19 of 33, or 19 of 31 depending ) :

Quote:
Although the initial interferometer is not sensitive to the gravity gradient due to the motion of a person walking in proximity of a test mass, advanced interferometers are projected to be sensitive to this noise source to a distance of 10 m


so the concern is not negligible! :-)

Quote:
Or is there more of a concern about where they are tunneling?...


Well, there was an issue with bats only a short while ago.

Quote:

Perhaps something else far too obvious and 'silly'... Perhaps try reducing all the detector dependant feedbacks gain in proportion to the laser intensity?... (Or is that supposed to have been done already?)

Or is the problem elsewhere and dependent directly on the small effects of heating from the laser beam? (Or anything of anything else!?)

Yes, yes, yes and yes! :-)

The behaviour is quite different for acquiring lock compared to maintaining lock ( see the feedback discussion earlier ). My impression/guess is now that the envelope of performance is being pushed, then what may have been subtle issues earlier are now assuming magnitude. This is not any degree of failure in the project, but an inevitable consequence of the approach to quantum limits. This has never been done before! Without being critical of anyone, it may simply be the novelty of the situation. Which is why I personally find this fascinating. :-)

Possibly a close reasonable everyday analogy would be motorbike racing. The effort to get into the top 100 qualifiers, say, does not really qualitatively relate to the effort to get into the top 10. So you tweak the bike, slightly change your approach to a given corner, account for the weather etc. It's notable that certain teams are at a disadvantage as they have no prior data on a particular track for instance. The devil is in the detail, and slight changes have non-linear results. And dumb luck of course. Our Phillip Island course produces bird strikes on a regular basis.

Actually the other terrific thing about LIGO and E@H is openness rather than occult stuff. I'm personally so pleased that we have such access to detail. Just imagine if one could have had a similiar level of involvement and timely disclosure during the time of the moon shots! Without sounding too corny, it is a bit like jotting down notes while looking over Einstein's shoulder! :-)

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

Here's some thinking from

Here's some thinking from Hanford on the IFO's behaviour :

Quote:

General Thought: Why The NS/NS Inspiral Range Drifts In The Long Run
[Science segment 520 has been running smoothly at 14W with a range of ~17Mpc for over 2 hours now...]

You'll notice from the attached pdf that the Range in the long run follows some sort of wave-like envelope curve. Some people think that this is just traffic on the roads, but I don't think that explains it.

It's intriguing because it seems to flow independent of the state, operator, power, or otherwise...in a 24-hour cycle. Not only that, traffic, as measured by the seismographs, doesn't seem to follow wave-like cycle through the night, but Range definitely does!

The closest channel I found to this pattern is TEMP05, the outside air temperature, which obviously follows a 24-hour pattern. If you look at the last page of my pdf, you'll see how close the patterns are. It seems as if TEMP05 leads Range by about 6 hours. Notice also that the upward slope of TEMP05 and range is much steeper than their range. Notice that their peaks are sharp and troughs are rough.

Questions:
1) Does LLO follow the same phenomenon? I'm still trying to get DataViewer to connect to their server.
2) Is the cause just some other daily 24-hour channel that resembles a wave? Like what?
3) Does this mean that as temperature drops in the winter our Range will drop as well?
4) Does seismic follow a similar pattern? I don't think so (look at pdf) but maybe I just didn't zoom in far enough or eliminate enough outliers.
5) Should I try a (sine) regression?


On reading this I immediately thought : air temperature down -> air pressure down -> air density up. But I can't as yet think of any consistent reasoning to explain any specific pattern. The pdf mentioned does vaguely show the range lagging the temperature by about 3-4 hours. The regression comment is to do curve fitting to characterise the trends better.

Asymmetric contraction of the ground?

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

RE: Add in some occult

Quote:
Add in some occult secret society trying all manner of obscure political and physical and occult tricks to try to secretly sabotage the tune up all to keep the secrecy of their great secret...


Well you could really whip up all the paranoid/conspiracy crews out there with the claim that LIGO is hiding data showing detection of the black hole at LHC! :-) :-)

I shouldn't goad. Some twit will probably believe it ..... *

At Livingston they are discussing reflective memory, which is neat as I've never understood it well:

Quote:

how reflected memory works
Based on some properties we observed, I think this is how the reflected memory system works:

1. Each computer on the reflected memory loop has its own chunk of memory dedicated to 'reflected memory'

2. When a computer reads from reflected memory, it reads from its own memory

3. When a computer writes to reflected memory, it sends out a little message around the reflected memory loop with the memory address and value to be written. As each machine on the loop receives this message, it writes that value to the appropriate address in its own memory, and then passes the message along to the next computer in the loop.

So, as long as all the computers' memories (1) start in the same state, and (2) remain in communication, then all the reflected memories will contain the same values.

However, if communication is interrupted, for example if a particular machine is put on 'bypass', then some machines will not get the 'write' messages, and then the computers on the reflected memory loop will in general disagree about what values exist at those addresses.

This explains the old story about 'bad data in the RFM' causing problems sometimes. The solution is to 'flush' the RFM, by having each machine re-write the values it is responsible for, so that the correct values get written to each machine's private copy of the reflected memory.

EDIT: The authoritative answer is at http://www.gefanuc.com/products/1288


So if the data becomes incoherent within the group bad stuff happens .... for control of a real time device especially.

At Hanford it's all been rather seismic :-)

Quote:
RE: Seismic activity in Severnaya Zemlya. Danny points out that this is where the Space Weapons Control Centre is located. Looking at the satellite feed, two Migs approach the complex and a blinding flash knocks out the satellite. It appears that the GoldenEye weapon has been deployed there


Cheers, Mike.

( edit ) * - mind you Tyrannosaur Canyon wasn't too bad a read, although it became a bit obtuse at the end. I think they are applying the Agatha Christie approach to SciFi : where the revealed 'solution' bears no relation to anything mentioned beforehand! :-)

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

As regards the

As regards the temperature/range 'correlation' at Hanford :

Quote:

NSNS/Inspiral Range Long Term Drift Revisited
..... I made a scatterplot comparing temperature to range, which as you can see (page 1 of PDF) does form a correlation, I think, once you remove the heavy outliers which are when Range=0. I'm still trying to figure out how to do that in ligoDV (suggestions?).

Peter Murray (Glasgow) asked me to look into the possibility that maybe range is higher at night because that's when we add power, from 8W to 14W. I looked at this in page 2 of the PDF. You can see that the steps from 8W to 14W (black dots) are too discrete to account for a wave-like pattern in Range (red dots). The bottom graph also shows curving when power is constant that this relationship isn't very strong.


The pdf is here. The scatterplot is a bit of a splatterplot.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6534
Credit: 284700169
RAC: 113726

Bunny update for Hanford

Bunny update for Hanford :

Quote:
19:47:25 UTC Lost lock, were not quite in science mode then. Reason: Spike in seismic around the LVEA. This is clear on FOM2. (Probably traffic, but I did spot the bunnies prowling at the picnic tables out back: See attached).


So, the plot thickens .... and I do hope they are not radioactive! They are getting some decent lock periods at good range with 14W .....

You can also see that the range is up/down/up/down .... at Livingston though, prompting the following comment :

Quote:
If you define interferometer velocity by inspiral range over lock time, then LLO is the fastest GW detector in the world.


perhaps due to the Output Mode Cleaner :

Quote:

1000 Took a look at the alignment, arms were off a little. IFO does lock but it doesn't make it through the script. Breaking at random spots so I believe it's due to the high microseism.

1117 Train arriving. When the IFO finally made it to the lock the OMC stage of the script I noticed the QPD3 and QPD4 SUMs were near zero so I started trying to get the light on the OMC QPDs. Will continue after the train departs.

1240 Still working on the OMC QPDs, I had the light back on the QPDs but lost lock before I found the carrier, the next the lock the light was off of the OMC QPDs again.


This is happening at 7W laser power. Mind you it's a busy time world wide for seismic activity and the Livingston site is on 'wobblier' ground than Hanford. But lock loss doesn't always evidently correlate with the seismic sensors.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.