A second level trigger (requiring hits in all 3 wire chambers in a sector)
has been implemented in hardware and software. The level 1 trigger
(whatever we choose to download to the CAMAC trigger) hasn't changed. The
ADCs and TDCs are still gated/started/stopped with the L1T from the
XMLU/Trigger Supervisor. The wire chamber decision takes longer, and so it
forms a second level trigger to either read out an L1T or reject it (and
hardware clear the ADCs and TDCs).
The software consists of changes to the FASTBUS and Trigger Supervisor
(TS) readout lists to enable 2nd-level triggers and fast clears of the
digitizers. These new readout lists are incorporated in a new CODA
configuration, physl2t.
The hardware consists in part of special boards to take hit wire
information from the back of the WC TDCs and form logical ORs and ANDs, and
then an OR of the two sectors. Groups of 15 wires may be jumpered to
"always TRUE" (for example, if you have to turn HV off because a cell won't
stay on).
The logic signal from the WC boards, the Level 1 Accept from the TS, and
the WC TDC Stop are used in NIM logic to create level 2 Pass and Fail
signals for the TS. On L2P, the TS tells the FASTBUS ROCs to read out the
event. On L2F there is no readout; the TS issues a Clear signal which we
send to the back of the FASTBUS crates to clear the digitizers. All of the
NIM logic and level translation is done in rack 3, in the NIMbin above the
scaler crate.
The level 1 and level 2 Accept signals from the TS are counted in the 11th
scaler module, labelled TSL1A and TSL2A. I believe that BDCCT/DCCT should
still be an appropriate measure of the experiment live time.
We should still be able to use the old level 1 trigger by selecting the
CODA physrun configuration. No hardware changes should be necessary. At
one time, I had to disconnect the fast clear cables (blue for the left
sector, yellow for the right sector) from the back of the FASTBUS crates,
but that no longer appears to be necessary. If physrun is the
configuration, then TSL1A and TSL2A will count the same. The only way to
know how much the L2T cuts down the event rate is to run with the physl2t
configuration.
Run 3444 is a cosmic TOF singles run with the second level trigger. Run
3447 is with the old first level trigger. The L2T takes the PHYS0 rate of
about 850Hz and cuts it down to about 20Hz (and there are 2 hot wires, so
the real 3-chamber-hit rate is even lower). For an L1T, the TS is dead for
about 5.2usec. For an L2T, the TS is dead for about 800usec. nsed looks
reasonable for run 3444. No attempt has been made yet to achieve a
"cosmics yield" and verify that the L2T doesn't throw out good events, but
things looked reasonable on the scope.
Karen
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:30 EST