ITMX tripped 26 June on Stage1 Actuators at 1209 utc. A 6.4EQ in Kyrgystan at 1117 is a good candidate.

Attachment 1 shows 5 seconds of the MASTER_dof_DRIVE--the output from the model; and, the three horizontal inputs to the model control loops.  Note the H2 drive going fuzzy just after T=29sec.  Any output from the model pushing H2 would also drive H1 and H3 which don't see any drive.  The bottom plot shows the RZ and X motion increasing at the same time while Y motion is unaffected.  A drive only on H2 alone would produce little if any Y motion (on this platform.)

On attachment 2, the horizontal coil monitor currents and voltages are plotted on the middle and lower graphs.  Different time scale with 30 seconds here but T=0 is the same.  Note at T=20 the H2 current does an ugly reverse and hump for a few seconds and then goes back to looking normal.  The H2 Drive and the plant response does not see any of this so it would seem this does not actually get to the Actuator.  The monitors do appear pretty normal until just before T=30 when it goes flat.  This when the H2 DRIVE has shot to narnia.  Note the WD does not trip until almost T=32 after the 8192 model cycles.

While the coil driver monitors show a different behavior here compared to the 13 July EQ trip discussed in 28567, the same arguements hold: 1) no model output only drives H2.  2) H2 only actuation would be sensed most by RZ and X; and, mostly not sensed by Y.  3) The H2 current monitor behavior several seconds before the badness must be suspect.  The voltage trend of H2 is trending higher at the time the extreme RZ and X response starts but it isn't a crazy number especially when compared to H1 voltage at the same time.  The 13 July event sees the H2 current ramping to high levels as the RZ and X sensors  respond.