Hello, All:
This shift was a study in frustration.
The wire apparently "sees" the beam halo in layers. With no Cerenkov
counters working for the trigger, we were forced to use the
elastic_tof_front4_back4 trigger (to hopefully help enforce the ep
coincidences). However, this trigger has a very low trigger rate (about
5-10Hz or so). The process for generating continual high trigger rates is
to continually move the wire in during a fill's lifetime as the halo dies
off. However, due to the halo layering effect with the wire, you cannot
really tell where the next layer is (the only places where the trigger
rate is at all reasonable for data taking) until you are already on top of
the layer.
However, being on top of the layer is bad, because it, for some reason,
causes a very high momentary trigger rate (>200Hz) consisting of
"fake" events (i.e. events not originating within the target) when the
layer is reached. This means that each time a layer is reached, many
"fake" events are written nearly instantly, followed by a relatively
longer sequence of lower rate events (within which the "good" events
originating from the target would be). Within 10-15seconds, the good
trigger rate has died down, and the wire needs to be moved in again. The
process then repeats.
The problem is that, due to the high "fake" event rates, not many events
originating from the target are recorded. (In addition to the high "fake"
rates at each layer, there is also a constant 1Hz event rate present even
with no wire target; all these add together to mean that, in any run, only
a small percentage of the recorded events will actually originate within
the target.)
I spent the first four hours of the night trying to come up with a scheme
to continually keep the trigger rate high yet at the same time avoid the
massive "fake" event dump when each layer is reached. I took many runs
(2034 - 2047) mostly consisting of one fill each; unfortunately, nearly
all of them are littered with "fake" events. Run 2047 looks good (in the
sense that nearly all the events had less than 200 hits in them) as long
as you only look at the first 3000 events or so (it has 3600 build events
on it; I hit a layer during the last 600 events).
It should be mentionned that this beam halo layering effect is a
relatively new phenonemon. Three days ago, it did not exist. Whether the
beam or the wire or both have changed since then, I do not know.
Around 4:30, the wire appeared to not be moving even though the software
said it was. Bob and Ernie were called in. I think they are trying to
determine if a motor needs to be switched or not. Anyway, as of 8:00, the
wire target is not working.
Aaron
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:28 EST