Документ взят из кэша поисковой машины. Адрес оригинального документа : http://zebu.uoregon.edu/~uochep/talks/talks03/G030112-00.pdf Дата изменения: Fri Aug 1 01:18:56 2003 Дата индексирования: Tue Oct 2 11:24:08 2012 Кодировка: Поисковые слова: п п п п п п п п п п п р п р п р п р п р п р п р п р п р п р п р п р п р п р п

Introduction to Hardware Signal Injections
Peter Shawhan, Isabel Leonor, Szabi MАrka
(LIGO/Caltech, U. of Oregon, LIGO/Caltech)

LSC Meeting March 19, 2003

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Uses of Hardware Signal Injections
Complete end-to-end test of detector and analysis
Verify that detector does not distort si gnal in unexpected ways Verify that search algorithm can detect signals ! Check that correct parameters are recovered Checks both calibration and search algorithm

Study coupling of GW signal into auxiliary channels
To evaluate "safety" of veto conditions

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Signal Injection Architecture
Implemented October 2001

Application

awg
Client library
ethernet

real-time process
VME CPU

Excitation channel
e.g. LSC-ETMX_EXC

Control-room workstation

Application passes waveform data to client library
Also specifies starting time for injection

Client library sends data to awg in 1-second chunks
Several seconds of buffering

awg adds a data value to digital controller every 1/16384 sec
Synchronized with GPS clock
LSC Meeting, 19 March 2003 Shawhan, Leonor, Marka LIGO-G030112-00-D


Client Library
a.k.a. "SIStr" library (Signal Injection Stream) Simple application interface
Open stream (specify channel and starting time) Append waveform data to stream Pass any number of data values at a time Make as many "append" calls as necessary Close stream (sends all data, then releases excitation channel)

Library handles buffering and data transmission timing
Application just has to provide data as fast as it can Arbitrarily long waveforms can be sent

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Limitations
Waveform data must account for transfer function of actuator
i.e. weight desired displacement by f
2

Can't inject arbitrarily large signals
Limited by dynamic range of actuator Dewhitening affects available dynamic range as a function of frequency

Can't have two processes injecting into same excitation point
e.g. can't have calibration lines and simulated astrophysical waveform both going into LSC-ETMX_EXC

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Client Programs
awgstream
Reads waveform data from an ASCII file Can specify start time and an overall scale factor

multiawgstream
Injects a sequence of waveforms at specified times

Others could be written
Could calculate waveform on the fly

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


,QMHFWLQJ 0XOWLSOH :DYHIRUPV
multiawgstream
c client software for injecting multiple waveforms in series c uses SIStr library by P. Shawhan and D. S igg c user can run multiple multiawgstream clients to inject into
multiple channel s s imultaneously (e.g., inject simultaneuosly
into H1/H2:LSC-ETMX_EXC, H1/H2:LSC-ETMY_EXC)

c typing "multiawgstream" by itself will give usage
From a control room machine, type: multiawgstream
[-q] [-d]

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


1HHGFRQILJXUDWLRQILOH([DPSOH
# Hanford S2 burst configuration file for H1:LSC-ETMX_EXC # Specify waveform data files and aliases wf wf wf wf wf f f f f f i i i i i l l l l l e e e e e /m /m /m /m /m ypat ypat ypat ypat ypat h/ h/ h/ h/ h/ wf wf wf wf wf s s s s s g100Q9. g235Q9. g361Q9. g554Q9. g850Q9. dat dat dat dat dat wf wf wf wf wf s s s s s g100Q9 g235Q9 g361Q9 g554Q9 g850Q9

# Specify signal aliases, scale factors, and time offsets wf wf wf wf wf wf wf wf s s s s s s s s g100Q9 g100Q9 g235Q9 g361Q9 g554Q9 g850Q9 g850Q9 g850Q9 0.5353 1.5604 1.1372 3.3150 9.6635 28.1701 40.4747 100.5366 21.0000 41.0000 61.0000 81.0000 101.0000 121.0000 141.0000 161.0000

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Injections During S1
Correlated noise (for stochastic group)
Two sets of files injected simultaneously into L1+H2, with a few amplitudes One set of files simultaneously injected into H1+H2 (one amplitude)

Inspiral chirps
Into ETMX_EXC in some cases, DARM_CTRL_EXC in others Two inspiral waveforms, various amplitudes

Burst waveforms
"Sine-Gaussians" (various frequencies) injected into L1+H2

Shortcomings (due to limited time)
Calibration lines were not present at time of injections Limited number of waveforms were injected Injections were not done systematically into all interferometers
LSC Meeting, 19 March 2003 Shawhan, Leonor, Marka LIGO-G030112-00-D


Injections Before S2
Tried to do things more systematically than for S1
Time for injections was allocated as part of the run plan Try to cover all interferometers (but susceptible to locking problems) Keep calibration lines going during injections Took a while to shake down how to do this Most successful approach: leave calibration lines on ETMX, inject waveforms into ETMY

Correlated (stochastic) noise
A few sets of files, a few amplitudes H1+H2 (no calibration lines) H1+L1 (calibration lines on ETMX, but with odd amplitudes)

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Injections Before S2
Inspirals
Several different waveforms, including some high-mass ones Injected into ETMX, ETMX-ETMY, ETMX+ETMY Some very loud into ETMX Had calibration lines on DARM

Bursts
Sine-Gaussians at various frequencies Lots of different polarizations and amplitudes Some loud, though limited by dynamic range of actuation (at least at some frequencies) Had calibration lines on DARM_CTRL

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D


Injections During S2
"Intra-run" injection sequence
18 inspirals + 24 bursts, injected into ETMY 30-minute sequence, done several times so far

Plan for a night of additional studies near end of S2 run
Night of April 9-10 Will need to set priorities and make plan in next ~2 weeks

LSC Meeting, 19 March 2003

Shawhan, Leonor, Marka

LIGO-G030112-00-D