Triggering

Like most modern colliders, HERA produces a large total interaction rate which exceeds the readout rate and data storage capabilities of the H1 and ZEUS detectors and necessitates the use of a sophisticated trigger system to select the physics events of interest. The trigger systems, as described in section 4 are complex multi-channel and multi-layer trigger and filter systems, which are optimized to maximize the statistics of selected physics processes, while suppressing beam backgrounds and downscaling the high-rate inclusive $ep$-scattering processes with no particular final state signatures.

The majority of heavy quark events is produced in the kinematic region of small transverse momenta $p_t$ and small photon virtualities $Q^2$. This kinematic region is of particular interest as the various scale variables $Q^2$, $p_t$ and $m_c$ or $m_b$ are of small and/or similar size. The total cross section in this region is dominated by processes with light quarks. Experimentally, particle identification is difficult and beam backgrounds are abundant. Triggering events with heavy quarks is particularly challenging as the effective rate of events with identifiable final state signatures is suppressed by the relatively small heavy quark production cross sections, the subsequent branching ratios, and the experimental acceptance limitations.

Typical trigger conditions used to collect events with heavy quarks implement a number of conditions in order to achieve reasonable purity of the triggered data samples and acceptably low rates. The efficiencies for heavy flavour triggers strongly depend on the physics channel (i.e.event selection) and range from $\sim 50\%$ for low-multiplicity events (e.g.$J/\psi$) in photoproduction to $\sim 90\%$ in electroproduction.

In both experiments H1 and ZEUS, triggering relies heavily on the evaluation of information from the tracking and calorimeter devices. The ZEUS trigger algorithms are more calorimeter-based, exploiting the excellent time resolution of the calorimeter, while that of H1 emphasizes tracking algorithms for reconstruction of the interaction vertex. Several detector components are used for the suppression of backgrounds from cosmic rays or beam gas interactions and for the identification of events with particular final states:

• The fast calorimeters (those with time resolution $\sim$ 1 ns) are used to select $ep$ events based on the arrival time of the scattered electron signal (H1) or all final state particles (ZEUS). Out-of-time backgrounds are further suppressed by veto-conditions using coincidences of signals in scintillator counters situated along the beam pipe. At ZEUS, the shaping, sampling, and pipelining algorithms permit the reconstruction of shower times with respect to the bunch crossings with a resolution of better than 1 ns. The timing information provides essential rejection against upstream beam-gas interactions.

• The H1 first level track trigger makes use of proportional chambers and drift chambers to determine the number of tracks, the event vertex position and the time. The central proportional chambers are used for fast reconstruction of the position of the interaction vertex in $z$ and of the time of the interaction. The main purpose of this trigger is the suppression of proton beam backgrounds which produce tracks from vertices outside the interaction region. Furthermore, the trigger is used to estimate the event multiplicities. The trigger information is based on combinations of pad signals of the proportional chambers from which track directions are inferred using look-up tables. 16 histograms (one for each $\phi$ sector) containing the $z$-positions of the tracks extrapolated to the beam axis are combined in a $z$-vertex histogram which is used to determine the position of the vertex in $z$. The $z$-vertex trigger provides trigger elements to the L1 system, encoding significances and multiplicities of the vertex information, and more detailed information to the L2 system. The H1 drift chamber trigger finds charged tracks in the $r$-$\phi$ projection. Drift time patterns from digitized hits in several layers (with wires parallel to the beam axis) are compared to predefined masks to determine kinematic properties of the tracks. The charge of the tracks can be measured and several transverse momentum thresholds, configurable in value, are used to classify tracks and to count track multiplicities. The system is optimized for the measurement of tracks from the interaction region and thus suppresses backgrounds due to beam-wall interactions or cosmic ray particles.

  At ZEUS, the drift chamber trigger alone determines the number of tracks and whether they originate from the interaction region. This is done by applying lookup tables to two dimensional projections of the $r$ and $z$ coordinates of the hits in the central and forward track detectors. The rate reduction is obtained by the rejection of beam backgrounds and the downscaling of inclusive electron-proton scattering events with no particular final state signature for events with $Q^2< 20$ GeV$^2$.

• The muon triggers use signals from the inner drift chambers and signals in the instrumented iron of the central muon detector. The H1 muon detector trigger is segmented into several modules. Coincidence of hits in several layers of the same module lead to a positive trigger signal from the muon systems. At ZEUS a muon track candidate in the central drift chamber with one or more hits in the muon chambers can be validated by energy in the calorimeter above a threshold of 460 MeV.