new RCG standard filter modules should start OFF
From bugzilla ticket #1134
Jeffrey Kissel (LIGO - Hanford Observatory) 2019-01-11 08:06:52 PST
Change request for RCG - Please make new filter modules start in the OFF configuration. Currently, new filters start as pass-throughs, which means that signals start flowing before the operators can adjust the setting. Bad control signals flowing to hardware can be a serious issue.
Jeffrey Kissel (LIGO - Hanford Observatory) 2019-01-11 08:07:42 PST
This bug is a copy of Integration Issue and ECR Tracker Ticket 12044 https://services.ligo-la.caltech.edu/FRS/show_bug.cgi?id=12044
Keith Thorne (LIGO - Livingston Observatory) 2019-01-11 08:10:06 PST
This may negatively impact all the test stand users, who will be very confused when new filters do not work.
Jonathan Hanks (LIGO - Hanford Observatory) 2019-01-11 11:37:14 PST
One way to do this might be to have the epics sequencer/SDF system look at all uninitialized channels. If a channel happens to be a gain for a filter we could have SDF force it to 0 on startup.
This behavior could be configurable via command line argument, environment, parameter in the .cmd file, ....
Jeffrey Kissel (LIGO - Hanford Observatory) 2019-01-14 11:15:43 PST
In reply to Comment 2: (1) Test stands should not drive the need for this change. We have lost significant and precious commission time to this "feature." (2) Any confusion can easily be rectified by paying attention -- both to this bug, the RCG release notes, this ECR/IIET resolution, and/or the work permit to change the RCG installation at each site. Model changers are super-users that are clued in to and paying attention all of this already. If somehow they don't find out, then the biggest penalty is them having to go in and turn the filter on in their desired configuration. A very low cost comparatively. (3) If a human adds a new filter, or collection of filters, to a system, that same human should be responsible for turning it on, and insuring its functionality. This should not be done automatically by computer.
Acceptance Test
New Filter Module in Existing Model
- Add a new filter module to an already existing model
- Rebuild/Install/Restart
- Open the SDF table, and verify that the EPICS VALUE of the new filter module is DC and the IN,OT bit is not set, also verify the gain is 0.0
Fresh Install
- After the fresh install model has been started
- Verify that only the DC bit is set, and that the gain is 0.0 for all filter modules and in the generated safe.snap