aLIGO LHO Logbook

H2 SUS

jeffrey.garcia@LIGO.ORG - posted 23:38, Tuesday 04 October 2011 - last comment - 08:02, Wednesday 05 October 2011(1506)

FMY M1 stage damping loops closed

The damping loops for the M1 stage of the FMY on the BSC8-ISI were closed today with the gains tuned. Now that the BIO medms are in place, the 'H2:SUS-FMY_M1_CTRL' value was set to "0xaaa" to enable the test coils. The 'H2:SUS-FMY_M2_CTRL' channel was set to "0xfff" for the M2 stage.

The new damping gains for each DoF are:
H2:SUS-FMY_M1_DAMP_L_GAIN = -1.0
H2:SUS-FMY_M1_DAMP_T_GAIN = -10.0
H2:SUS-FMY_M1_DAMP_V_GAIN = -1.0
H2:SUS-FMY_M1_DAMP_R_GAIN = -0.3
H2:SUS-FMY_M1_DAMP_P_GAIN = -0.05
H2:SUS-FMY_M1_DAMP_Y_GAIN = -0.1

The "burtgooey" gui was used to save a burt snapshot from the request file:
'/opt/rtcds/lho/h2/target/h2susfmy/h2susfmyepics/autoBurt.req'

A burt snapshot of the new configuration is located in:
'/opt/rtcds/lho/h2/userapps/release/sus/h2/burtfiles/fmy_dampfilter_111004.snap'

Comments related to this report

jeffrey.kissel@LIGO.ORG - 08:02, Wednesday 05 October 2011 (1507)

Link

Other notes:
This suspension's EPICS values had been entered using the test stand's
/ligo/svncommon/SusSVN/sus/trunk/BSFM/X1/Common/Scripts/Prepare_BSFM_06272011.m,
However, this script contained a few bugs, and the there are a few differences between the test stand and the production electronics that required a few changes.

(1) AA / AI sign flip. The test stand has an older version of the AA/AI filter board, which has a DC gain of -1. This sign had been compensated for on the test stand in the OSEMINF banks (for the AA sign), and the COILOUTF banks (for the AI sign) (see LLO aLOG 1672, and subsequent comments, for full explanation of all the sign flips in a test stand's signal chain). On the production units, which have later versions of these boards where the DC gain is +1, we no longer need to compensate for this sign change. So, the sign of gains in both the OSEMINF and COILOUTF bank must be flipped from what is hard coded in the teststand's Prepare script. HOWEVER, the COILOUTF bank still must compensate for the magnet polarities that differ between OSEM basis degrees of freedom, so the signs of each gain in these filter banks should now be (in F1 F2 F3 LF RT SD order),
OSEMINF = [+ + + + + +]
COILOUTF = [+ + - + - -] (corresponding to [N N S N S S])

(2) OSEM2EUL / EUL2OSEM matrices. A while back, I'd caught a bug (1321) in the function the generates/calculates the matrix elements for these matrices
/ligo/svncommon/SusSVN/sus/trunk/BSFM/Common/MatlabTools/make_susBSFM_projections.m
-- the pitch lever arm had mistakenly been hard-coded in as 0.55 m, when it should have been 0.055 m. These matrix elements, which account for the rotational degrees of freedom's lever arms to preserve units through the basis transformation, were therfore off by a factor of ten. The Prepare script, which merely calls this function, was run before this bug was caught. No biggie, just called the make*_projections function and installed the elements "manually" using
/ligo/svncommon/SusSVN/sus/trunk/Common/MatlabTools/fill_matrix_values.m.