DfiLpa-StatsAscStatistics.hh File Reference

The ACD Software Counter Statistics. More...

#include <DFI/Dfi-Acd.hh>

Classes
class	DfiLpaStats::Asc::SummaryV0
class	DfiLpaStats::Asc::CnoHistogramV0
class	DfiLpaStats::Asc::CnosV0
class	DfiLpaStats::Asc::TileHistogramV0
class	DfiLpaStats::Asc::TilesV0
class	DfiLpaStats::Asc::StatisticsV0
Namespaces
namespace	DfiLpaStats::Asc::CnoHistogramV0Enum
Enumerations
enum	DfiLpaStats::Asc::CnoHistogramV0Enum::Bin {   DfiLpaStats::Asc::CnoHistogramV0Enum::A = 1 << 0,   DfiLpaStats::Asc::CnoHistogramV0Enum::B = 1 << 1,   DfiLpaStats::Asc::CnoHistogramV0Enum::AB = A | B,   DfiLpaStats::Asc::CnoHistogramV0Enum::BinMin = 0,   DfiLpaStats::Asc::CnoHistogramV0Enum::BinMax = AB,   DfiLpaStats::Asc::CnoHistogramV0Enum::BinCnt = 4 }
enum	Discriminator {   DfiLpaStats::Asc::TileHistogramV0Enum::PA = 1 << 0,   DfiLpaStats::Asc::TileHistogramV0Enum::PB = 1 << 1,   DfiLpaStats::Asc::TileHistogramV0Enum::HA = 1 << 2,   DfiLpaStats::Asc::TileHistogramV0Enum::HB = 1 << 3,   DfiLpaStats::Asc::TileHistogramV0Enum::V = 1 << 4,   DfiLpaStats::Asc::TileHistogramV0Enum::DiscriminatorMin = PA,   DfiLpaStats::Asc::TileHistogramV0Enum::DiscriminatorMax = V,   DfiLpaStats::Asc::TileHistogramV0Enum::DiscriminatorCnt = 5 }

---

Detailed Description

The ACD Software Counter Statistics.

Author:

JJRussell - russell@slac.stanford.edu

   CVS $Id: DfiLpa-StatsAscStatistics.hh,v 1.3 2011/03/29 20:58:45 russell Exp $

A set of ACD Software Counter statistics consists of 3 pieces

• An overall set of summary counters. This is a 3x3 array of counters that keep track of the number of times the two data structures, the GEM veto counters and the ACD contribution used as a source for the data are
  • Present
  • Missing
  • In error There are no known ways that the GEM contribution can be missing. so this set of 3 counters is almost certainly 0. A missing ACD contribution is possible, but very unlikely. The most probable reason for a missing ACD contribution would be either an error in the transport fabric (moving the data from the AEM to the LCB or an error in the event structure which resulted in a failure to locate the ACD contribution. Either the GEM or the ACD contributions could be in error, for example, a parity error during the transport.

• A set of counters for the 12-CNO bits in the GEM contribution. These are captured in 6 histograms, representing the 6 board pairs, each with (technically) 4 entries, representing the correlated counts of the A and B sides of the board.
  • 0 Neither the A or B side counts
  • 1 Only the A side counts
  • 2 Only the B side counts
  • 3 Both the A and B side count One expects that the majority of counts occur in entries 0 and 3, that is either neither side counts or both side counts. Entries in the A-side only or B-side only could indicate
  • Differences in the signal strengths. This,is most likely, is due to differences in the physical pick and transport of the signal. There is no real action, other than noting it, that can be taken.
  • Imbalanced discriminator settings. Small differences are to be expected, but a large difference may be due to improper settings of one or both of the discriminators.
  • A noisy tile on one side
  • A dead tile on one side. The only way to distinguish these effects is likely to examine the pulse height distributions from actual events. In extreme cases, these pulse height distributions may have to made subject to various preconditions, such as demanding a track point at the board group in question.