|
|
Trigger Background Remediation
Last updated July 20, 2000.
Contents: [Explanation] [Offline Analysis] [Trigger Rates] [Fast Monitoring] [Deadtime]
[Jan 99 Workshop]
Recent Update
Summary of Jan/04 Background Run Info
Explanation
Since backgrounds represent an essential part of all Trigger work, this page is devoted to answering questions raised by the BABAR Background Remediation group.
BaBar Background Remediation web page
Summary: The trigger background is dominantly caused by inelastic scattering of lost particles in the beam wall resulting in hadronic showers. The contribution to the trigger rate from the HER is a factor of two larger than from the LER presumably due to the higher energy resulting in harder interactions. But the dynamics of the interactions follow similar patterns: the dominant background scatters occur +- 20cm at a flunch of the beam tube. This is at the beginning of the B1 magnet. The HER dominates the west side, the LER the east side. At -20cm the HER in addition sees a cooling maningfold. A secondary background contribution arises inside B1 at several masks. The LER distribution shows two secondary peaks at +40 and +60cm, the HER a single peak at -50cm. These peaks result from interactions in the opposite wall side than the central peaks, the HER hits here the east side, the LER the west side. This is also the expected behavior from the PEP II design.
Offline Analysis of L1 Pass-through Events
Most Recent Analysis of L1 Pass-through Events in Collision Runs in January 2000:
Track vertices, directions, charge and pt
[Run 10990 (ps)]
Page 1: Track multiplicity, d0 and z0 at L3 Trigger Level and offline reconstructed
Page 2: Two track sample, Track vertices derived from circle fits of fully-reconstructed tracks
Page 3: One track sample kinematics
Page 4: Separate plots of East/West Vertices
For comparison see Collision Run from August 1999 [Run 8266 (ps)]
The background distributions have not changed between August 1999 and January 2000. All distributions are HER dominated and look very similar to the single HER beam data (see below). The only contribution from LER can be seen on page 2, the little East-peak in the x distribution. Also the polar angular distribution shows the dominance of the HER to the L1 trigger background.
Most Recent Analysis of L1 Pass-through Events in Single Beam Runs in February 2000:
Track vertices, directions, charge and pt
[HER (ps)]
Page 1: Track multiplicity, d0 and z0 at L3 Trigger Level and offline reconstructed
Page 2: One track sample, track intersection with beam tube, x, y and z projections
Page 3: One track kinematics, polar angle, charge and pt
Page 4: Closer look to track intersection: separate plots for East/West
Page 5: (More than) Two track sample, Track vertices derived from circle fits of fully-reconstructed tracks
Page 6: Kinematics of vertex tracks
Page 7: Separate plots of East/West Vertices
A list of the background runs from February 1 can be found [here].
More PEP details [Log of Witold Kozanecki].
See also L1 Pass-through Events in Single Beam Runs from August 1999: Track vertices, directions, charge and pt
[LER (ps)][HER (ps)]
A list of the background runs from August 17/18 can be found [here].
Plots from Background-triggered Events:
Trigger Event Display [LER (ps)][more LER (ps)]
[HER (ps)][more HER (ps)]
My Multi-Event Display (shows track fit parameters) [LER (pdf)][HER (pdf)]
Offline Analysis of Cyclic Trigger Events
Most Recent Analysis of Cyclic Trigger Events in Collision Runs from November 1999:
Track number, pt, nHit, d0, z0 and correlation with physics trigger lines
[Random Trigger (ps)]
Plots from Cyclic Trigger Events:
Trigger Event Display (in Collision) [1 (ps)][2 (ps)]
My Multi-Event Display (in Single Beams) [LER (pdf)][HER (pdf)]
Trigger Rates
Most Recent Rate Measurements in Collision Runs (June 2000)
[L1 Trigger Rate versus HER current and versus time(ps)]
Page 1: The L1 rate for fixed HER currents (in multiples of 100 mA) as a
function of time, starting from February 1, 2000. We can devide the 4 months
period into 3 categories. (1) The trigger rate is reduced due to hardware
changes aiming to cope with the high background levels. (2) The HER current
is constantly pushed higher, but the trigger rate stays roughly constant.
This is due to scrubbing of the vacuum chamber. A scrubbing effect of
10% reduction in the rate per week is seen for HER currents of 700 mA and
above. (3) The HER vaccum leakage puts an administrative limit on the HER
current of 700 mA. In this period the LER current is pushed higher and we
observe a slight increase in the trigger rate.
Page 2: The L1 rate as a function of the HER current,
starting from February 1, 2000. Each color block corresponds
to one week. Two effects of opposite direction influence the trigger
rate: the rate increase due to higher currents and the scrubbing
reduction, that are similar in magnitude but opposite in sign and lead to
an almost constant trigger rate over time.
Page 3: The L1 rate as a function of the LER current, starting
from April 8, 2000 after the HER vacuum leakage. The trigger rate
increases linearly with the LER current from 550 to 750 Hz as the LER
current is raised from 700 to 1200 mA.
Page 4: Trigger Development from June, 8. No conclusions are derived
at this time.
Most Recent Background Rate Measurements in Single Beam Runs (July 2000)
[L1 Trigger Rate versus LER current (ps)]
Page 1: Total L1 Trigger Rate versus LER current. The fitted slope is 130 Hz
per A.
[L1 Trigger Rate versus HER current (ps)]
Page 1: Total L1 Trigger Rate versus HER current in 3 different settings.
Blue = HER only, Red = HER with 1100 mA LER out of collision, Black = HER
with 1100 mA LER in collision. The HER only slope is 360 Hz per A. With
both beams present but out of collision we have the same slope of 360 Hz per A
only shifted by an offset of 140 Hz as expected from the LER contribution
(see first ps-file). In collision the slope is significantly higher, 640 Hz
per A, where the difference shows the rate increase due to luminosity.
Page 2: The difference in the trigger rate between colliding and non-colliding
is shown as a function of the luminosity. The points fit to a linear curve.
The slope is 70 Hz per 10^33 of luminosity.
Extrapolation to higher currents (July 2000)
The main good news is that we now have reached a trigger background
that scales linearly with the machine currents. This is very remarkable
and reflects the good PEP II vacuum. It is for the first time in July
2000 that we have seen a linear scaling whereas for previous times we
always have seen a quadratic term (see below measurements from February
2000 and August 1999). This is confirmed by the long time observation of
the trigger rate in collision runs (see scrubbing described above). In
April 2000 the HER quadratic term was suppressed by a factor of 10 compared
to August 1999 and in July 2000 has virtually disappeared. An other
interesting observation is that most of the luminosity-related trigger
rate is physics. We find 70 Hz per 10^33 of luminosity from which 52 Hz
(bhabha's) + 7 Hz (muon, tau and uds, c, b pairs) = 60 Hz are known
physics channels. On top of this is a rather poorly known rate of two
photon physics, leaving only little room of less than 10 Hz per 10^33 for
beam-beam tail background. This is again very good news that the beam-beam
background is small and it is mostly physics that is responsible for the
luminosity-related trigger rate.
We now can parametrize the trigger rate in a simple formula:
L1 / Hz = 130 (cosmics) + 130 x LER / A + 360 x HER / A + 70 x L / 10^33
In the Seeman-Model of machine evolution we get:
L1 Trigger Rate Extrapolation
[(ascii)][(ps)]
Provided the good vacuum conditions we reached in July 2000 the single-beam
contributions to the L1 trigger rate, even in the year 2004, will not
exceed 1 kHz. By then the dominant contribution will come from the
luminosity. For instance, for 10 times design luminosity of 3 x 10^34 the
rate would increase by 2.1 kHz. This is unacceptable high. But most of the
rate increase comes from Bhabha's and the problem can be solved by prescaling
the L1 Bhabha line 1Y. We can summarize, the trigger is operational at 3 x
10^34 of luminosity in the present hardware configuration but needs a backup
option for poor vacuum conditions in the begin after a vacuum break.
Background Rate Measurements in Single Beam Runs (February 2000)
[L1 Trigger Rate versus HER current (ps)]
Page 1: Total L1 Trigger Rate. The first plot shows the rate increase
for single HER beam on February 1, 2000 with respect to August 18, 1999.
The HER background has worsend by almost a factor of two. In addition
there is a HER beam instability starting at 500mA. The point above
500mA cannot be fitted to the parabola below 500mA. The second plot
shows again the rate as a function of the HER current, but now in the
presence of 800 mA LER beam interleaved by 4 buckets from the HER. The
upper curve is the measurement, the lower one is the expectation from
single LER and HER beams (measured separately and then added together).
The difference is the beam-beam cross-term. Again, this cross-term has
increased significantly. In August 1999, the cross-term was less than
10%. Although the beams were interleaved, the HER increases now the gas
pressure in the LER pipe and vice versa. In addition we obersved a
drift of 10% in the trigger rate in a time interval of 10 min. This
indicates a temperture related outgasing problem.
L1 Line specific Rate Analysis by Jeff Booth (February 2000)
[Individuell L1 Trigger Line Rates (ps)]
Page 1: LER and HER rates in comparison between August 1999 and February 2000.
Page 2: Ratio of LER and HER rates between August 1999 and February 2000.
Page 3: Same as page 1, but last bin shows total L1 rate.
Page 4: Same as page 2, but last bin shows total L1 rate.
Page 5: LER and HER contributions to L1 Trigger Line Rates in August 1999 (top)
and February 2000 (bottom).
Page 6: Same as page 5, but last bin shows total L1 rate.
Background Rate Measurements from August 1999
[L1 Trigger Rates versus LER and HER current (ps)]
Page 1: Total L1 Trigger Rate
Page 2-7 : Individual Trigger Lines 1B, 1M, 2M, D2, D2*+ and 3M.
D2*+ is the essential trigger for dimuon and bhabha events, 3M is an open
trigger for hadronic events.
Page 8: Trigger rate contribution from LER, HER individually, the LER-HER cross term and the collision term at LER=800mA, HER=300mA. The HER dominates the background. Relative contributions are: 50% HER, 25% LER and 25% sum of beam-beam terms. The EMC is more sensitive to the HER, whereas the DCH sees about equally LER and HER. In particular the higher multiplicity of the HER background is problematic for the physics triggers. Thats why the HER contribution increases for the higher multiplicity trigger lines.
Page 9: Extrapolation to LER=1200mA and HER=450mA. Basicly the limit of the current trigger setup. The total rate reaches 1 kHz.
Page 10: HER depandance from IP steering angle
Event Size
Study of the Event Size in Background Runs by V. Serbo
[BaBar event size versus LER and HER current (html)]
Contributions are shown separately for all subdetectors and for different trigger lines (all logged events, L1 Pass-through events and cyclic triggers).
Previous Rate Estimates
BaBar Note #490 (S. Petrak, A. McKemey) summarizing the background rate measurements from the February '99 PEP II Commissioning Phase
["Results from the Crystal Commissioning Experiment for the PEP II Background Conditions" (ps)]
The 1M Trigger Rate in June '99 is a factor of 200 lower than the predicted
rate based on single crystal measurements in February '99. The uncertainty of
the February '99 prediction is mainly due to the extrapolation from one single
unshielded crystal to the entire calorimeter. This crystal was in a position
of the present EMC endcap. Occupancy measurements with bunchcrossing triggers
indicate a factor of 3.5 higher exposure of endcap crystals as compared to
barrel crystals. When correcting for the factor 3.5 still a factor of 60
discrepancy remains. This factor can be understood as due to a substantial
improvment of the PEP II beam conditions and a positive shielding effect of
the IFR doors on the Forward Endcap.
A brief Collaboration meeting talk summarizing Level 1 and Level 3 trigger rate studies done in Spring 1998. October 29, 1998, Level 1 and 3 rates talk (F. Kral) in "Backgrounds" mini-plenary [talk (scanned pdf)]
Fast Monitoring of backgrounds in Level 1
Trigger Rate & Dead Time Monitoring
Besides the total L1 rate and dead time there are two Trigger Lines,
1B (line 22) and 1M (line 23) that are used for backgrond monitoring.
Similar as for other background signals there is a duality how these
signals can be displayed. On BaBar side you find the 1B and 1M signals in
the lower left corner of the EPICS Background panel. The total L1 rate
and dead time can be viewed in a strip graph by clicking on the "Global
Strip Tool" button in the FCT panel. The Epics names for 1B and 1M are
CEN:ACC:TRG:1B:CALC and CEN:ACC:TRG:1M:CALC. Alternatively the trigger
channels are connected to MCC. They are accessible through SCP and archived
in the MCC database. From the SCP Index Panel, go to the BaBar Panel, and
then BaBar Analog Ch104 = 1M, Ch105 = 1B. For this you need an MCC account.
Deadtime in DAQ
is monitored in Epics. Follow instructions in Trigger Rate & Dead Time Monitoring
January 1999 workshop
Here are some notes from the January 13-15, 1999, Background Remediation workshop.
January 1999 workshop notes from Trigger (html talk)
Back to Trigger home page
Back to Trigger operations home page
Back to Operations home page
sibylle@slac.stanford.edu |
|
