Home » Blogs » kocolosk's blog

BBC Vertex

Updated on Thu, 2007-10-18 12:13. Originally created by kocolosk on 2007-10-18 11:04.

This is a study of 2005 data conducted in March 2006.  Ported to Drupal from MIT Athena in October 2007

Goal: Quantify the relationship between the z-vertex position and the bbc timebin for each event.

Procedure: Plot z-vertex as a function of trigger and bbc timebin. Also, plot distributions of bbc timebins for each run to examine stability. Exclude runs<=6119039 as a result. Fit each vertex distribution with a gaussian and extract mean, sigma.  Plots are linked at the bottom of the page.  See in particular page 8 of run_plots.pdf, which shows the change from 8 bit to 4 bit onlineTimeDifference values.

Timebin 12 had zero counts for each trigger. Summaries of the means and sigmas:

m 4 5 6 7 8 9 10 11 12
mb 101.7 124.1 56.3 22.72 -7.835 -38.89 -80.48 -135.6 -
ht1 81.09 111.8 50.75 16.88 -13.49 -44.57 -89.36 -182.8 -
ht2 79.86 107.2 48.93 15.76 -14.27 -45.17 -89.76 -176.7 -
jp1 101.3 139.3 54.8 17.68 -12.95 -44.05 -92.4 -176 -
jp2 99.49 128.5 51.58 15.77 -14.29 -45.5 -94.1 -192.2 -

 

s 4 5 6 7 8 9 10 11 12
mb 68.85 67.48 40.4 34.26 33.31 35.91 49.32 76.29 -
ht1 68.99 70.9 43.06 34.71 32.49 34.86 47.42 79.98 -
ht2 67.64 70.13 43.06 34.76 32.55 34.98 47.26 78.22 -
jp1 76.49 79.85 45.44 35.6 32.56 35.34 49.69 78.1 -
jp2 76.18 78.01 45.73 35.79 32.87 35.68 49.37 81.21 -


Conclusions: Hank started the timebin lookup table on day 119, so this explains the continuous distributions from earlier runs. In theory one could just use integer division by 32 to get binned results before this date, but it would be good to make sure that no other changes were made; e.g., it looks like the distributions are also tighter before day 119.  Examining the vertex distributions of the different timebins suggests that using timebins {6,7,8,9} corresponds roughly to a 60 cm vertex cut. A given timebin in this range has a resolution of 30-45 cm.

»