I was working on shaking the HAM3 ISI to see the effect of potential clipping on the PR2 scraper baffle (with the idea that we were then going to move the spot on PR2 and see how DARM changed), but I was saturating the H3 actuator and the ISI's watchdog timed out and tripped off.
I had just increased the excitation by a factor of 2, to 10 counts going to H1:ISI-HAM3_ISO_Y_EXC, and that last factor of 2 is what started saturating the actuator. Since we were able to see some subtle effects in DARM with 5 counts, we'll try using that when we next get locked. We may be interested in a conversation with SEI folks about whether it is safe to override the actuator saturation watchdog during this type of test, in order to more clearly see the effect in DARM, but we'll save that for another day.
Sheila pointed out that it would be useful if verbal alarms would tell us when we are saturating a seismic actuator (similar to how it tells us if we are saturating a suspension).
Attached are screenshots of the amount of motion above ambient, as well as that this last factor of 2 increase in actuation is what caused the saturation monitor to start accumulating, as well as of the excitation settings. This DTT template is a copy of what was in the Noise Budget git space for HAM5, with the exception that the noise budget excitation value is 0.6 counts, and I wasn't really seeing an effect in DARM until I got to 5 counts.
Here's also the HAM3 watchdog plot, again showing that it was my actuation that caused the trip.
There is also the suspoint drive infrastructure that we've added to all of the ISI models, allowing you to drive the euler basis motion for a particular suspensions top stage suspension point. Not sure if that will help for this measurement or not. For each ISI there are filter modules called ISI-CHAMBER_SUSPOINT_DRIVE_L/T/V/R/P/Y, giving the necessary testpoints, EXC channels for those fms are stored in the daq at 2048. You will need to populate the downstream eul2cart epics matrix to get the correct translation to the ISI cartesian drives. Output from the eul2cart matrix is then summed into the error point of the cart basis isolation loops. The matl code for writing the coefficients is really simple, for PR2, on HAM3 it would be:
addpath /ligo/svncommon/SeiSVN/seismic/Common/MatlabTools/
cd /opt/rtcds/userapps/release/isc/common/projections
load('ISI2SUS_projection_file.mat')
sus={'pr2'}
for ii =1:length(sus)
fill_matrix_values(['H1:ISI-', upper(ISI2SUSprojections.h1.(sus{ii}).chamber), '_SUSPOINT_EUL2CART'],ISI2SUSprojections.h1.(sus{ii}).EUL2CART)
pause(.1)
end
This might fail on a couple of the coefficients, so it might be a good idea to run a couple times. Also, there's only 1 6-dof matrix per ISI, so if drives on multiple suspensions are needed on one table, you will need to populate the matrix for each suspension. Probably also best to zero out all of the matrix elements first, then write with the above code.