- oleg's home page
- Posts
- 2024
- 2023
- February (1)
- 2022
- 2021
- 2020
- 2019
- 2018
- 2017
- 2016
- December (2)
- November (1)
- October (2)
- August (1)
- July (1)
- June (3)
- May (2)
- April (3)
- March (4)
- January (1)
- 2015
- 2014
- December (2)
- October (2)
- September (1)
- August (1)
- July (3)
- June (1)
- May (3)
- April (2)
- March (2)
- February (1)
- January (1)
- 2013
- My blog
- Post new blog entry
- All blogs
FMS Run 15 Analysis: BBC vertex calibration (II)
(A) Primary vertex selection
For the FMS triggers, the usual selection of the highest ranked primary vertex is not good enough.
We typically have only one or two tracks with positive ranks to begin with (plus a few with negative rank).
The plot below shows the primary vertex and BBC TAC difference for different thresholds on the rank ( in steps of 1000, starting at 1000000).
For rank>1002000, the peak in the track vertex has gone away and the distribution is nicely Gaussian.
Correlation between primary track vertex and BBC TAC differecence (for whole Run 15 data)
There still is a difference in the slope from central to further upstream (white and gray lines).
(B) Slewing correction (simple)
While the primary vertex distribution for the rank>1002000 cut is nicely Gaussian, the BBC TAC difference still shows a small bump on the left shoulder.
Looking at different ranges of AdcSum in East and West, this seems to be coming from the slewing in BBC tubes.
Plots below show the TAC difference for ranges in AdcSumEast (left), AdcSumWest (right), and both (middle) (in intervalls of 500), starting from black (0-500) following through the rainbow blue, cyan, green, yellow, orange, red, magenta to gray).
The slewing is large at AdcSum<2000 and then flattens out. The effect is opposite for East and West, and it cancels between both.
The distributions are nicely Gaussian in the fit ranges; all spectra are normalized to the peak height (after the fit).
The slewing correction is described by a two dimensional function (assuming the correlation between x and y are negligible):
F(x,y) = A0+ ax/(x0+x) + ay/(y0+y)
x: AdcSumEast
y: AdcSumWest
A0 = 4164.9
ax = -508155
x0 = 1559.6
ay = 364896
y0 = 954.2
(C) Proper slewing correction
Proper slewing correction should be based on single tubes, i.e. earliest TAC hit and the respective ADC signal height.
This will need the st_physics MuDst data; st_fms statistics are not enough.
For comparison:
primary vertex width is about 50 cm
BBC vertex width is about 86 cm before simple slewing correction
BBC vertex width is about 77 cm after simple slewing correction
Groups:
- oleg's blog
- Login or register to post comments