User name R. Michaels
Log entry time 21:17:25 on September27,2000
Entry number 48429
keyword=DAQ/Trigger changes, Aug-Sep
Here I document some changes during the Aug-Sep 2000
down period which affect DAQ and trigger.
DETECTOR SWAP --
This affects a lot of software, among other things. Since
the detector package "electron" was moved to the right
spectrometer, its now called the "right" arm. There will
be some software which has the old notatation, at least for
awhile, so remember the correspondence
Electron = E = e = arm 1 = trigger 1 = Right arm = R
Hadron = H = h = arm 2 = trigger 3 = Left arm = L
For ESPACE it is more complicated because of magnets.
The old method of connecting to HV and other RS232
ports has been changed. Now we use "portserver" devices.
One can still use the old "vt100" style terminals, but it
is recommended to use the PCs instead. Whichever terminal,
one uses a command like
telnet hatsv4 2005
to connect to port 5 on the portserver "hatsv4". The mapping
of channels, and usage instructions, are at ~adaq/doc/portserver.doc
and are posted on several walls.
TRIGGER Supervisor on R-Arm
Since e97111 wants to run the R-arm as the slow spectrometer
it was necessary to put the trigger supervisor there, as well as
the coincidence electronics. This affects scalers, etc. The full
list of triggers are listed on page 8 of the R-arm xscaler display.
In the raw data they appear as the 8th scaler bank, a 32-channel
scaler with header 0xceb70020. The channels are 1-5 for triggers
1-5 etc. Most old cabling remains valid, but I didn't fan out the
triggers to the all old scaler locations, e.g. the L-arm doesn't have T5.
In addition to the measurement scheme during N-Delta experiment
(there was an halog about that sometime around July 5), we have
added one other paddle (#3). So, paddles 1 and 3 are pulsed at 10 Hz
(fixed rate) simultaneously on each spectrometer. The deadtime is
measured in scalers by counting the logical AND of the trigger with the
pulser. The pulser can be turned on and off. From the adaq account
on any adaq computer (any platform) type "rpulser on" or "rpulser off".
The scaler channels are
L-arm: 5th (0xabc4) scaler, 22nd channel=pulser, 23rd=pulser.AND.trigger
R-arm: 8th (0xceb7) scaler, 17th channel=pulser, 18th=pulser.AND.trigger
The deadtime is measured after the 2nd MLU but there are a few modules
between it and the trigger supervisor whose deadtime we neglect. This
should not matter because: 1) The rates after the 2nd MLU are low, and
2) The previous method (looking at tagged events, unfolding DAQ deadtime),
is still available as a cross check and is complete.