Virgo Logbook

Two new paths of the TCS_MAIN node have been prepared, each one related to one different measurement regarding the CH/DAS actuators ("Compensation" and "Centering"); the first part of the measurement, though, is in common between the two.

The two paths start with the common part from DOWN, then they separate at one point; a number of new DQ_META channels are defined and set. An overall summary of all the DQ_META_TCS channels (also those already existing) will be provided later in this entry.

Both paths rely on the same [COMPENSATION_MEAS] section of the TCS_MAIN.ini configuration file, with most of parameters overlapping.

The code is not online yet but it is ready, it will be deployed in a dedicated and planned time slot soon.

Description of the new paths of TCS_MAIN:

Common Part: initialization and Central Heating

COMPENSATION_MEAS_INIT (index = 35, non-requestable): here several parameters are read: which HWS to use (hence which plate will be interested), the HWS configuration file (used only to check its existence before waiting two hours to acknowledge that), and the cooldown_time. Then, all the flip mirrors will be put UP to block all beams; then, the cooldown_time will be waited before moving on.
COMPENSATION_MEAS_LOCK (index = 36): this is the steady state which will be requested by ITF_LOCK, in order to perform the operations described in the previous one. Concurrently, ITF_LOCK will keep the LOCKED_ARMS_IR configuration, and notify the same timeout (but the blocking timeout will be the one of TCS_MAIN).
COMPENSATION_MEAS_CH (index = 38, non-requestable): here the selected HWS will be initialized, the reference measurement will be performed, its outcome on disk verified, then the live measurement will be launched; the Automation will wait for cool_meas_time seconds (configurable default is 10 minutes), then the flip mirrors for the Central Heating will be put DOWN and the measurement will continue for a much longer warm_meas_time amount of seconds (default 2h) while the CH thermalizes back. This state will be requested from ITF_LOCK after it will have requested the MISALIGNED_ALL_BUT_{NI,WI} state to ARMS_LOCK; given that the SDB1 drift control is usually engaged directly by ITF_LOCK and not at the lower level of ARMS_LOCK, it guarantees it will not be engaged (which is what we want).

From COMPENSATION_MEAS_CH two paths can be walked.

Compensation Path: measurement of the relative compensation between the CH and the DAS.

COMPENSATION_MEAS_DAS (index = 39, non-requestable): here we switch back DOWN the flip mirrors of all the DAS actuators (beams will pass through), then the ongoing HWS acquisition will be continued for another warm_meas_time amount of seconds (default 2h).
COMPENSATION_MEAS_DONE (index = 40): steady state, reached once the measurement is finished. At this point all actuators are active and thermalized, and the lock acquisition can be attempted already.

Centering Path: measurement of the relative alignment between CH and DAS.

CENTERING_MEAS_INIT (index = 45): this is a steady state (where nothing happens), which is needed for state requests from ITF_LOCK; its point is to make TCS_MAIN go from COMPENSATION_MEAS_CH (end of common path) back to COMPENSATION_MEAS_INIT, in order to flip UP again all flip mirrors, while ITF_LOCK relocks the arm cavities and stays there for (again) a cooldown_time amount of seconds, while everything thermalizes again.
CENTERING_MEAS_DAS (index = 48, non-requestable): this is completely equivalent to COMPENSATION_MEAS_CH, but instead of the CH, once the cool_meas_time seconds of acquisition will have passed, only the relevant DAS_OUT flip mirror will be brought DOWN, then another warm_meas_time seconds of acquisition will be performed.
CENTERING_MEAS_DONE (index = 50): steady state, reached at the end of the measurement; here nothing happens with the exception that all flip mirrors will be moved DOWN, thus re-enabling all TCS beams on the CPs; however, in this case the thermalization will still need to happen after reaching this state.

Both paths then end up in DOWN.

As mentioned, there are several DQ_META_TCS channels that the automation sets to 0 or 1 depending not only on the automation state, but also on general status or measurements ongoing; they are all set to 0 both in INIT and in DOWN, and they are (all with DQ_META_TCS_ prefix):

FAULT: this is set to 1 when there is something wrong, usually in the configuration of a measurement, in order not to acknowledge that a few hours later; it is set to 1 while going to the equivalently-purposed TCS_FAULT Metatron state;
DAS_CALI_ON: this is set to 1 when the old, almost unused, DAS power calibration path was used (it relies on the PyTCS process);
HWS_{DET,INJ}_ON: it is set to 1 each time the corresponding HWS is turned on/used;
COMPENS_MEAS_{NI,WI}_CH_ON: it is set to 1 in the common path of the new automation, so when the HWS is actually acquiring in Single Bounce with everything off but the CH;
COMPENS_MEAS_{NI,WI}_DAS_ON: it is set to 1 in the Compensation path of the new measurements, where all the DASes are back on with the HWS acquiring;
CENTER_MEAS_{NI,WI}_ON: it is set to 1 in the Centering path of the new automation, where only one relevant DAS_OUT actuator is shining the CP, with the HWS acquiring.

ITF_LOCK has been modified in order to accomodate and use the new states of TCS_MAIN; its own new states are all non-requestable, and must be selected from the all menu. In the following, the ones in bold are the ones to be requested in order to perfom the full measurements. No predecessors are required and they can be requested from DOWN:

COMPENSATION_MEAS_COOLDOWN (index = 235): here we request TCS_MAIN to go to COMPENSATION_MEAS_LOCK, i.e. we flip UP all the TCS actuators, and we keep the arms locked while the ITF thermalizes down;
COMPENSATION_MEAS_READY (index = 236): steady state, reached once the process has finished;
COMPENSATION_MEAS_RUN (index = 239): here we request the full Compensation path, i.e. directly COMPENSATION_MEAS_DONE; ITF_LOCK will stay here (in Single Bounce) for cool_meas_time+2*warm_meas_time (all off + CH on + all on) seconds; it will not be blocking, as this will be enforced by TCS_MAIN not reaching its final target state before such an amount of time;
COMPENSATION_MEAS_DONE (index = 240): steady state, reached once the process has finished;
CENTERING_MEAS_CH_ON (index = 241): this is almost identical, in logic, to COMPENSATION_MEAS_RUN, but it will only request half of the path, i.e. the end of the common path, COMPENSATION_MEAS_CH; for this reason, it will wait for only cool_meas_time+warm_meas_time seconds (cold + CH) while in Single Bounce;
CENTERING_MEAS_CH_DONE (index = 242): steady state, reached once the process has finished;
CENTERING_MEAS_RELOCKING (index = 244): here we move on the other path, the Centering one, by asking TCS_MAIN to go to CENTERING_MEAS_INIT, which implies passing through COMPENSATION_MEAS_INIT again, while the arm cavities are relocked;
CENTERING_MEAS_RELOCKED (index = 245): steady state, reached once the process has finished;
CENTERING_MEAS_DAS_ON (index = 248): very similar to CENTERING_MEAS_CH_ON, where we do the second half of the measurement in Single Bounce, with only the DAS_OUT shining the CP.
CENTERING_MEAS_DONE (index = 250): steady state, reached once the process has finished.

The new states layouts of TCS_MAIN and ITF_LOCK are attached.