Displaying report 1-1 of 1.
Reports until 22:01, Friday 30 January 2015
H1 SEI (DetChar, SYS)
jeffrey.kissel@LIGO.ORG - posted 22:01, Friday 30 January 2015 - last comment - 10:36, Saturday 31 January 2015(16391)
H1 ETMX BRS-correct Sensor Correction Trips ETMX-ISI
J. Kissel, R. Adhikari, S. Dwyer,

By happenstance calling for another reason (see LHO aLOG 16390), Rana and Sheila informed me of woes at ETMX: a sudden trip, some further trip, followed by more trips of the watchdog. Unsure of what was going on, they had mashed a bunch of buttons in attempt to fix the problem. 

After some remote MEDM'ing, I could tell that the BRS-rotation-corrected, GND X STS signal had gone bonkers, which was being used as the X -direction Sensor Correction signal since Krishna had resurrected the BRS earlier this afternoon (see LHO aLOG 16380). 

I've since switched the STS2CART input matrix to use the normal X-direction STS signal for sensor correction (changed the STS2CART matrix to use STS B for X input to sensor correction instead of STS C). I'm not sure what's gone wrong with the BRS (too difficult to diagnose remotely), we'll figure that out later.

Further, Sheila and I used the SDF system to restore all of the button mashing they'd done. Now only the expected differences from "nominal" remain -- my switch of the STS2CART input matrix, and Krishna's turning on the outputs of the Rotation and Torque correction paths for the STS. Again, the latter no longer has any impact because this GND super sensor output is no longer being used. I second Hugh's love of the new SDF system!
Images attached to this report
Comments related to this report
sheila.dwyer@LIGO.ORG - 03:28, Saturday 31 January 2015 (16395)

this is the first trip

Images attached to this comment
krishna.venkateswara@LIGO.ORG - 10:36, Saturday 31 January 2015 (16401)

I've attached the BRS output trend for the last ~day and the problem is clearly visible at ~4:00 UTC. As I'd mentioned in the previous alog, BRS was outside of its nominal range. When the BRS gets close to the edge of it's range, the subtraction between the reference and the main pattern doesn't work right and produces spikes in the output, which is what seems to have happened. It was a mistake to leave it running in the sensor correction loop in this state and I apologize for it.

On Tuesday, during maintenance, I will change the set-point of the balance physically to bring it in range.

Images attached to this comment
Displaying report 1-1 of 1.