Timing the Bump
                      -------------------


                                                 Original Rev.  15-MAR-2006
                                                   Current Rev. 15-MAR-2006


This note is a description of how we will determine the arrival
time of the peak of the BLS signal "bump" in the ADF_2 cards.
This is a task that will be handled "in house" by MSU as compared
with determining the energy calibration which will require software
from another institution.

Timing the Bump  (TtB)  will require some beam time.  We will not
need the DAQ system to actually make a run for these dedicated
studies because all data is handled within the ADF_2 cards and the
L1 Cal Trig Trigger Control Computer  (L1_Cal_TCC).  During this
dedicated beam time all other DAQ system components must be quiet,
specifically the SMT system should not be injecting its readout noise
into the Calorimeter.


Philosophy

  - Separate tasks for collecting the data and for analyzing the data.

  - This is new territory and there will be noise in some channels
    so don't hide the raw data from the humans.

  - Signell's rule applies as usual.



Description of the TtB Data Collection

Positive and negative Eta Rings have different cable lengths and
must be handled separately.  Within a given Eta Ring the arrival
time of the bumps at all Phi's should be the same.  This is by design
in the cable plant.  Thus no effort will be put into automatically
separating out the arrival times of different Phi's.   


Modes:  Start running TtB  or  Stop running TtB


Assumptions about the system before starting TtB:

    standard  T7_Phy firmware has been configured into all cards

    all cards have been Initialized

    all channels have their Zero Energy Response DACs set
    for a Zero Energy ADC output of 50 decimal counts


What we tell TtB at the start:

    run for T minutes

    the eta range to run over

    threshold for each eta

    definition of a good bump
       >= M continuous ADC samples above 1/2 of the threshold
       <= N continuous ADC samples above 1/2 of the threshold
    this is a common definition to all eta's


What TtB writes into a file sorted by eta's:

    this is a hit for phi = 1:32

    value of the peak ADC sample

    number of continuous ADC samples that were over 1/2 of the threshold  

    The actual address in the Raw ADC Memory Block  where the peak
    ADC sample was found.   Coin toss if 2 ADC samples at peak.
    The TtB process on L1_Cal_TCC should read forward through the
    Raw ADC Memory Block and report the first  (earliest in time)
    peak that it finds.  The TtB process should not look for a second
    peak in that Raw ADC Memory Block.

    Offset from the memory block address holding the peak ADC sample
    to the memory block address where we expected the peak ADC sample
    to be.  Calculating this value requires knowledge of the run time
    parameter, "OffsetTick1RawAdcMonit",  and knowledge of the beam
    structure.

        Recall that when collecting this TtB data the Address
        Generators will stop at the end of a turn.  Collecting this
        data does not involved the lookback calculations, i.e.
        "OffsetTriggeredToTickAtCapture" needed for data that is
        captured in response to L1_Acpt with Collect Monitor Data.




Realtime Servicing of the ADF_2 Cards During a TtB Data Collection Run    

   The various Trigger Towers will report hits as they find them.
   The TtB process on L1_Cal_TCC must give equal service to
   different regions in eta,phi   e.g. not always start at +1,1.