[d2n-analysis-talk] Trigger Plots for Coin. and LHRS runs (& evtypebits description)

[Brad Sawatzky]

Hi Dave,

It's not clear from your description if the cuts in
THaDecData::TrigBits() have been updated or or not.  That is the
calibration I was mentioning.

The N'th bit of the DL.evtypebits data word is intended to tell you
whether the hardware trigger N initiated the CODA event you are looking
at.  The code that handles setting those bits is in THaDecData.C.  The
DL.evtypebits variable is constructed as follows:

  1 A copy of all hardware triggers is sent to a set of TDC channels.
    You (the user) use DecData class's database to assign each of these
    TDC channels a variable name of the type 'bit%d'.  The class
    instance is conventionally named something like 'DL' or 'DBB' so you
    get variables named 'DL.bit1', 'DL.bit2'...
  2 If the DecData class sees any hit in the the DL.bitX[] variable
    between 200--1500 then the X'th bit in DL.evtypebits is set.

Question/Issue 1:
  If the comments in db_DL.dat are to be trusted, the channels assigned
  to the bitX variables in step (1) are actually the trigger signals
  passed through the Trigger Supervisor /after prescaling/.  I am not
  convinced that that is correct -- please put it on your list for
  discussion at the next meeting (ie. remind me).

Question/Issue 2:
  That range in step (2) is pretty arbitrary and can cause problems if you
  have a high rate trigger that generates accidentals.

  In principle, if trigger X is what forms the level 1 accept (L1A) that
  actually triggers the TDCs and ADCs, then those events should form a
  self-timing spike in the DL.bitX histogram (the stop is the same signal
  as the start, only with some fixed hardware delay between them).  The
  cut in step (2) would ideally cut around that self-timing spike.  (In
  practice this can get a little complicated, particularly on coincidence
  runs where the TDC may be triggered off a retimed signal.)

  By histogramming the DL.bitX variables you can identify the self-timing
  spike that tells you trigger X actually formed the trigger for the event
  you are looking at.

You and Matt are to confirm that the cuts in step (2) make sense, and to
correct them if they do not.

-- Brad

On Mon, 26 Oct 2009, David Flay wrote:

> I've been looking at trigger plots for the coincidence elastic run 1229
> and an LHRS run 20718 (4-pass beam, p = 0.60 GeV)
> 
> The plots are attached in a tarball for the T1, T3 trigger (run 1229), and
> T3 trigger
> (LHRS 20718).  In each plot there are trigger cuts shown, utilized in the
> form:
> 
> (DL.evtypebits&(1<<3))==(1<<3) for T3, for instance.
> 
> The cuts are color coded and labeled on each plot.  For the coincidence
> run, the variable DL.t* was used, for the LHRS run, the variable DL.LT*
> was used (only DL.T3 and DL.T8 variables are valid for the LHRS run, while
> DL.t* has DL.t1...DL.t8 available).
> 
> Is there anything further you want Matt or I to do with the trigger plots?
>  I remember you mentioned something concerning a calibration?

-- 
Brad Sawatzky, PhD <brads at jlab.org>  -<>-  Jefferson Lab / Hall C / C111
  Ph: 757-269-5947 -<>- Pager: 757-584-5947 -<>- Fax: 757-269-7848
The most exciting phrase to hear in science, the one that heralds new
  discoveries, is not "Eureka!" but "That's funny..."   -- Isaac Asimov