New H1 Track Triggers

The H1 Collaboration has upgraded the trigger electronics (described in section 5.7) in order to be able to improve the signal to background ratio and to collect useful physics events while suppressing the triggering of events that are not of interest or are beam-related background.

The H1 Fast Track Trigger (FTT) [237] implements charged track finder and mass reconstruction algorithms in the first three levels of the H1 trigger scheme. For the charged particle track reconstruction at the first trigger level 1 (2.3 $\mu$ s) and level 2 (23 $\mu$ s), the FTT makes use of Field Programmable Gate Arrays (FPGA) and their embedded Content Addressable Memories (CAM). The track parameters are determined by comparing hit-patterns with predefined masks implemented in digital signal processors (DSP). The FTT can reconstruct up to 48 tracks which is sufficient for about 98% of the events of interest.

The FTT functionality is based on hit information in the central jet chambers CJC. In fig.56, the geometrical cell structure of both chambers is sketched. In the first step track segments are formed separately in four groups of three layers of wires each. A fast track segment linking, based on matches of track segments in $\kappa (\propto 1/p_t)$ and $\phi$ is completed within 2 $\mu$ s to provide level 1 trigger decisions based on charged track multiplicities and charged particle topologies for coarse cuts. The result is used by the second level FTT where the track segments are linked and re-fitted to better precision within 20 $\mu$ s including the determination of event quantities like a refined track multiplicity, momentum sums and invariant masses for low multiplicity events. Figure 57 shows the measured precision for FTT tracks at level 2. The resolutions are given here with respect to the full off-line reconstruction. The resolution in is approximately $2\%$ , approximately 2 mrad in $\phi$ and 75 mrad in $\theta$ .

The track parameters of the fitted tracks are sent to FTT L3 where a farm of commercial processor boards is used to perform a full search for particle resonances within 100 $\mu$ s. The L3 track information is either used directly or in combination with information from other trigger subsystems to generate a final L3 decision.

**Figure:** - **$\phi$** view of a charged particle track from the interaction point traversing the central drift chambers of the H1 experiment. The sense and cathode wires of the chambers are indicated.
$\begin{figure}\unitlength1cm \begin{picture}(10.,6.5) % put(0.,0.)\{ epsfig\{fig... ...put(2.,0.){\epsfig{figure=figs/ftt.eps,height=6.5cm}} \end{picture} \end{figure}$

**Figure:** Measured resolution of the FTT algorithm at second trigger level relative to the full off-line CJC reconstruction for a) for tracks with MeV and b) **$\phi$** for tracks with MeV.
$\begin{figure}\unitlength1cm \begin{picture}(10.,6.7) % put(0.,0.)\{ epsfig\{fig... ....7cm}} \put(2.,5.){\large a)} \put(10.,5.){\large b)} \end{picture} \end{figure}$

In addition to upgrading the drift chamber trigger, the H1 experiment has improved the robustness against severe background conditions by replacing its double layer central inner multi-wire proportional chamber by a chamber with five layers of pads with geometry projective to the -position of the interactions. The new detector has ten times more channels and, like the old proportional chamber, provides trigger decisions, based on a -vertex histogramming technique, at level 1, within a latency of 2.3 $\mu$ s.