With the loops closed I ran the $SCRIPTS/MC/WFS/senseWFSoutMATRX script and analysed the lockin outputs with $SCRIPTS/MC/WFS/matlab/wfsmatrix3.m. I had to edit both the setupWFSlockins and the sensWFSoutMATRX scripts because in the past we used to switch on / off the ASC filter bank GAINs on the MC suspensions to start / stop the lockin excitation. We cannot do this any more since these the WFS feedback signals have to get through these filters while the WFS loops are closed. So the current, more sensible, scheme is to set the appropriate elements to 1 / 0 in the C1IOO_LKIN_OUT_MTRX.
Note: The senseMCdecenter script will also have to be ammended in the same manner.
The lockin outputs measured are (reduceddata):
wfs1P wfs2P mc2tP wfs1Y wfs2Y mc2tY
MC1P -10.3694 7.0642 -10.2133 -0.1025 0.4653 -0.0000
MC2P 8.2838 21.5141 8.4102 -0.2215 0.0734 0.0000
MC3P 9.4804 6.0835 9.6346 -0.0080 0.0366 -0.0000
MC1Y -0.7339 -1.4498 -0.6175 -11.7502 -13.0480 0.0004
MC2Y 0.9004 0.6645 1.0554 25.6083 7.3399 -0.0046
MC3Y -0.2914 2.1573 -0.1829 -2.1130 14.3038 -0.0000
After inverting and normalising a subset of the above matrix ( done in the wfsmatrix3.m ) we obtain the following output matrix coefs:
Apart from a negative sign (introduced by the negative gains in the WFS servo filters ) these values are quite close to the actuation vectors determined in open loop.
I have plugged these values into the WFS output matrix. Will determine the open loop gain later when there arent so many people stomping around the MC.