The trigger supervisor has died. The problems Adrian and Tavi had last
night were due to no -2V in the trigger supervisor crate. Chris Brown
and I replaced the TS crate with a spare this morning. For a while the
TS worked, but it was flaky. Now it doesn't work at all. JLab is
FedExing another one, but it won't be here until Monday.
In the meantime, I tried to resurrect using the SFI front-panel to
trigger the FASTBUS readout, the way we did in the DTF:
* Made a new configuration for RunControl, commis, that doesn't use the
TS
* Made copies of the FASTBUS readout lists (in ~/commis/coda/crl),
commis_l_pri_nots.crl and commis_r_pri_nots.crl. I believe I made all
the right changes to trigger from the NIM 1 input of the SFI front
panel, instead of the auxiliary backplane connector. On node blast,
compiled these following the README in that directory.
* The "Run in progress" signal comes out of NIM 1 output of the SFI.
Put that in a quad coincidence module (above the scaler crate) along
with the physics trigger from the MLU.
* The "readout not in progress" comes out of NIM 2 output. That should
be delayed then used to stop a busy circuit (see CODA DAQ logbook 1 p.
140) in the counting bay. I'm not really worrying about that for now.
The NOT of "busy" is ANDed with (physics trig AND run) to make the
readout trigger.
* That readout trigger comes out of another quad coincidence section
above the scaler crate. It is sent back to the SFIs, NIM 1 input.
* Disconnected the FASTBUS-TS interface from the FASTBUS backplane for
each crate, just in case.
I see good triggers in the readout trigger cable. But the SFIs are not
responding to it. I can't figure out the problem. Perhaps the evening
shift can make it work. If not, we can't run FASTBUS events until the
replacement TS arrives.
Karen Dow
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST