CDS Maintenance Summary: Tuesday 17th September 2024

WP12091 Add new TEST_NOTIFY Guardian Node

TJ, Dave:

TJ created the new node and the new H1EDCU_GRD.ini. I updated H1EPICS_GRD.ini. DAQ+EDC restart needed.

WP12087 Add VACSTAT gauge channels to DAQ.

Dave:

I added the gauge metadata channels to H1EPICS_VACSTAT.ini. DAQ+EDC restart needed.

This failed due to channel name lengths exceeding 54 characters. This change was then backed out.

Verify PEM signal cabling to h1iopcdsh8 IO Chassis

Fil, Dave:

Fil verified the PEM cabling was as per h1pemh8. See Fil's alog for details.

New LIGO 28AO32 DAC driving ETMX L2 and ESD signals

Daniel, Richard, Marc, Fil, TJ, Tony, Dave, Erik, Jonathan, EJ

The 28AO32 DAC was wired to drive the SUS ETMX DAC channels which were being driven by its 20bit DACs. Existing AI chassis were used for these drives to remove one more change.

Two additional AI chassis were installed in this rack. One to keep the first 20bit DAC in a good state, the second to drive the h1susetmxpi DAC drives (last two channels of second 20bit DAC).

DAQ Restart

Jonathan, Dave:

The DAQ + EDC were restarted for the Guardian and VACSTAT changes. This was a very messy restart.

DC0 did not restart, it found that the new VACSTAT channel names were too long.

I reverted the change to H1EPICS_VACSTAT and we did a second restart of the DAQ 0-leg and EDC.

FW0 spontaneously restarted itself after only writing a few full frames.

When we were sure FW0 was stable again, we restarted the 1-leg with no issues.

The name-too-long issues should have been caught by my check_daq_channels_validity.py code. I subsequently found there was a bug in the code, and it was not checking the new DAQ INI files, rather the running version. This has been fixed and tested.