CDS Maintenance Summary: Tuesday 11th March 2025

WP12374 WP12370 ASC and SUS model changes.

Daniel, Jeff, Jonathan, Erik, TJ:

New models were installed for h1asc, h1susim, h1sushtts and h1suspr3. A DAQ restart was required.

WP12372 Add ISI HAM1 cards to h1seih16 IO Chassis

Fil, Marc, Erik, Dave, Jim, TJ:

h1seih16 was powered down. In its IO Chassis the bad A2 Adnaco backplane was replaced and the cards which had been moved to A3 were returned. Two additional ADCs and one additional BIO6464 were installed as per drawing (attached) which will be used by the new h1isiham1 model. Currently these cards have no external cables attached.

A new h1iopseih16 model was installed with these cards added. Note an additional 16bitDAC was not required because seismic ham models allocate a DAC per chamber, with the HEPI model using the lower 8 channels and the ISI model using the upper 8 channels. DAC-1 was already installed for h1hpiham1.

Note that this was one of two changes to the h1iopseih16 model (see SWWD section).

DAQ restart was required.

WP12373 Install new h1isiham1 Model

Jonathan, Erik, Jim, Dave:

The first version of the ISI HAM1 model was added to h1seih16. It is currently running with nothing connected to its ADC, DAC and BIO cards.

This is not something we do regularly, so we took the opportunity to document the steps needed. The list we have so far has been included as a spreadsheet.

The model was added to the DAQ. A related change to h1hpiham1 was also made. A DAQ restart was required for both.

Attachment shows the CDS Overvew MEDM h1seih16 section, now with a full complement of models.

WP12382 HAM1 SWWD

Jeff, Dave:

In preparation for HAM1 ISI install, we created a HAM1 Software Watchdog (SWWD) on h1iopsush2b and h1iopseih16. Note this is the second change to h1iopseih16 today.

There are two suspensions on HAM1; RM1 and RM2.

Despite its name, h1iopsush2b also controls ham1 suspensions. A SWWD part on h1iopsush2b reads the UL, LL, UR, LR raw ADC channels and calculates their integrated RMS values. If any value exceeds the trip level, the Dolphin IPC channel sent to h1iopseih16 is set to zero.

The IPC receiver on h1iopseih16, on receipt of a trigger signal, starts a DACKILL count-down timer. If the timer expires, it executes a DACKILL for the second DAC card, which drives HAM1 HEPI+ISI.

New h1iopsush2b and h1iopseih16 models were installed. The h1iopsush2b was done at the same time as the h1susim and h1sushtts models since these are the only models running on this front end. h1iopseih16 restart was done when this front end was powered back up after the IO Chassis work.

The SWWD MEDMs were updated with the new HAM1 system (attached)

Note: one complication is that up until now SUS SWWDs have been for larger suspensions with 6 OSEM channels (F1, F2, F3, LF, RT, SD). RM1,2 only have 4 (UL, LL, UR, LR) and so the two unused OSEM inputs have been grounded in the models. On the MEDM I just show ADC0_30, the channel prior to the duotone which is zero.

WP12358 Move Digital Video Cameras from h1digivideo2 to h1digivideo5

Patrick, Jonathan, Erik, TJ, Sheila, Dave:

We moved the six cameras which were on the last of the old servers (h1digivideo2) onto the new AMD/deb12 h1digivideo5 and powered down the old machine.

All was good until locking started, which is was discovered that for cameras which are copied to front end models the X,Y centroid EPICS channels now had offsets which the models could not handle.

For locking expediency it was decided to move the cameras back to the old server for now until the models could be configured to use to the new centroid values. Of the six models, five were moved back to h1digivideo2 (BS, ITMX_RED, ITMX_GREEN, ITMY_RED andITMY_GREEN). POP-AIR is not copied to FE models and so could stay on h1digivideo5.

At time of writing H1 has locked and been in OBSERVE for 50 minutes.

To "green up" the EDC I am running a dummy IOC on opslogin0 to serve the new channels for these cameras which were lost when we reverted them to the old software.

The camera MEDM shows which ones are still running on VID2 (attached).

DAQ Restarts:

Jonathan, Erik, Dave:

Due to the large number of changes we did two DAQ restarts today. The first was primarily for the ASC and SUS model changes and SWWD. The second was for the addition of the new h1isiham1 model and the camera move.

The EDC restart was due to new:

H1EPICS_DIGVIDEO.ini (at the time, moved 6 cameras to new server)

H1EPICS_DAQ.ini (add h1isiham1)

H1EPICS_FEC.ini (add h1isiham1)

H1EPICS_SDF.ini (add h1isiham1)

The camera server code on both h1digivideo4 and h1digivideo5 was updated from 0.1.16 to 0.1.17 to fix a bug in the periodic archiving of camera images (https://git.ligo.org/cds/software/cameras/pylon-camera-server/-/issues/6).

In addition Tony and Jonathan removed h1digivideo0 and h1digivideo1 from the racks as a cleanup job as part of the camera update WP.