jeffrey.kissel@LIGO.ORG - posted 11:51, Monday 19 November 2012 - last comment - 15:04, Tuesday 20 November 2012(4730)
H1 SUS ETMY Infrastructure Completion
I have filled out H1 SUS ETMY's new infrastructure, including (1) Filling in the ISI to SUS WIT and SUS to ISI OFFLOAD paths, installed as was described for H1 SUS PR2 in LHO aLOG 4553. Nothing new -- same exact filters -- just using the H2 SUS ETMY (yes, H2 -- the SUS hasn't moved yet) matrices from /opt/rtcds/userapps/release/isc/common/projections/ISI2SUS_projection_file.mat (2) Installing the OPTICALIGN alignment offset calibration gain, as described in LHO aLOG 4563. Here, for this and all QUADs, the gain is OPTICALIGN GAIN [cts/urad] = ( DAC [V/ct] * Coil Driver [A/V] * Coil/Magnet [N/A] * HSTS M1 to M3 [rad/N.m] * 1e9 [nrad/rad] )^-1 = ( (20/2^18) * 0.009943 * 1.694 * 0.033514 (for Pitch) * 1e9 )^-1 = 0.023219 [cts/nrad] for Pitch = ( (20/2^18) * 0.009943 * 1.694 * 0.015055 (for Pitch) * 1e9 )^-1 = 0.051689 [cts/nrad] for Yaw Note that I've gone with nanoradians instead of microradians (as was done for PR2), because the alignment offsets needed to align (at least the H2OAT) were already quite large (~75% of the M0 BOSEM range), so calibrating this gain into microradians (or any gain larger than 1, really) would have saturated the DAC. This provides for annoying large EPICS values, (retaining the -3027 and +5405 [ct] offsets needed for the H2OAT, means that these offsets are 58562 and 232783 [nrad]), but I think we can deal. I still need to adjust the OPTICALIGN screen to account for this difference (make the sliders have bigger default range and step size, as well as changing the legend to match). (3) With further quantitative assessment pending, I've installed Rana's 2k to 16k anti-imaging filter in the L3_ISCINF_P L3_ISCINF_Y filter banks, in order to filter any imaging noise that arises from the inter-process-communication between the ASC model (running at 2k) and the SUS model (running at 16k). I attach a bode plot of the filter as installed (again, identical to Rana's).
Non-image files attached to this report
Comments related to this report
As of this entry, I've updated and committed H1 SUS ETMY's safe.snap here,
${userapps}/sus/h1/burtfiles/etmy/h1susetmy_safe.snap
as well as saved and committed the updated foton filter here,
${userapps}/sus/h1/filterfiles/H1SUSETMY.txt.
For both, I've checked that they're obeying the new schema, in which
/opt/rtcds/lho/h1/target/h1susetmy/h1susetmyepics/burt/safe.snap
and
/opt/rtcds/lho/h1/chans/H1SUSETMY.txt
are soft links to the above userapps locations.
S. Aston, J. Kissel
Stuart was double checking my work (hooray!) and found several problems with my implementations of the OPTICALIGN offsets and gains.
(1) There's a typo in the labelling of the calculation (I wrote "for Pitch" twice, when one should be Yaw). The corrected statement is
"For this and all QUADs, the gain is
OPTICALIGN GAIN [cts/urad] = ( DAC [V/ct] * Coil Driver [A/V] * Coil/Magnet [N/A] * HSTS M1 to M3 [rad/N.m] * 1e9 [nrad/rad] )^-1
= ( (20/2^18) * 0.009943 * 1.694 * 0.033514 (for Pitch) * 1e9 )^-1
= 0.023219 [cts/nrad] for Pitch
= ( (20/2^18) * 0.009943 * 1.694 * 0.015055 (for Yaw) * 1e9 )^-1
= 0.051689 [cts/nrad] for Yaw
(2) In my haste, I'd installed these gains exactly backwards (0.023 for YAW and 0.052 for Pitch). In fact, even further, I had calculated the offsets to-be-restored with the gains exactly flipped as well. Whoops! I've now corrected the gains, and changed the stored offsets to
Pitch: -3027 [ct] / 0.023219 [cts/nrad] => -132302 [nrad]
Yaw: +5405 [ct] / 0.051689 [cts/nrad] => 104548 [nrad]
I'm Currently taking measurements, but I'll take a new safe.snap once I'm finished.
New safe.snap has been captured and committed to the userapps repo, as of version 3401: /opt/rtcds/userapps/release/sus/h1/burtfiles/etmy/h1susetmy_safe.snap