View source for DVCS-Hodoscope trigger system design 19-Oct-2007

I'm thinking the easiest test can be the done following way:

We can setup the system with the discriminators all tied to the same load
(i.e. use a fanout device where ultimately a single input can be fanned
out to all discriminators). We can then create test patterns by using a
pulser to drive the single input of the fanout to the discriminators. Then
we can create any cluster pattern by setting the discriminator thresholds
low so they're triggered by the pulser. Since the discriminator thresholds
are real-time progrmamable over VME I can have a program read hit patterns
from a file and setup the discriminators and then record for which events
the system finds a trigger.

This would test the trigger at the full-speed (though events won't be
generated at full speed they will be processed at full speed)...Of course
we'll need several fan-out units (15x 1->32 ECL fanout boards: 1->32->448)
and a bunch of LEMO cables (but you'll need those in the end anyway).

If this doesn't sound feasible we can elect to do a simpler setup, but it
would be wise to test the entire setup.

For the Hodoscope:
If all the detectors in one quadrant of the tower calorimeter can depend
on no more than 32 elements of the hodoscope (say for an 80 element
hodoscope, 20 elements will go to each quadrant of the calorimeter + 12
more for shared edges = 32 inputs). If this setup works, then we can use 2
additional V1495 boards. Both outputs from the hodoscope discriminators
would feed each extra V1495 board using 3x 32b input ports on each. Then
each V1495 (connected to Hodoscope discriminators) can feed 2 quadrants of
the tower calorimeter V1495 boards.

I can update the drawing showing this setup.

Ben