As the Tevatron luminosity increases sophisticated selections are required to be efficient in selecting rare events among a very huge background. To cope with this problem, CDF has pushed the offline calorimeter algorithm reconstruction resolution up to Level 2 and, when possible, even up to Level 1, increasing efficiency and, at the same time, keeping under control the rates. The CDF Run II Level 2 calorimeter trigger is implemented in hardware and is based on a simple algorithm that was used in Run I. This system has worked well for Run II at low luminosity. As the Tevatron instantaneous luminosity increases, the limitation due to this simple algorithm starts to become clear: some of the most important jet and MET (Missing ET) related triggers have large growth terms in cross section at higher luminosity. In this paper, we present an upgrade of the Level 2 Calorimeter system which makes the calorimeter trigger tower information available directly to a CPU allowing more sophisticated algorithms to be implemented in software. Both Level 2 jets and MET can be made nearly equivalent to offline quality, thus significantly improving the performance and flexibility of the jet and MET related triggers. However in order to fully take advantage of the new L2 triggering capabilities having at Level 1 the same L2 MET resolution is necessary. The new Level-1 MET resolution is calculated by dedicated hardware. This paper describes the design, the hardware and software implementation and the performance of the upgraded calorimeter trigger system both at Level 2 and Level 1.

We report on a search for pair production of first-generation scalar leptoquarks ($LQ$) in $p \bar{p}$ collisions at $\sqrt{s}$=1.96 TeV using an integrated luminosity of 203 $pb^{-1}$ collected at the Fermilab Tevatron collider by the CDF experiment. We observe no evidence for $LQ$ production in the topologies arising from $LQ \bar{LQ} \to eqeq$ and $LQ \bar{LQ} \to eq \nu q$, and derive 95% C.L. upper limits on the $LQ$ production cross section. %as a function of $\beta$, where $\beta$ is the branching fraction for $LQ \to eq$. The results are combined with those obtained from a separately reported CDF search in the topology arising from $LQ\bar{LQ} \to \nu q \nu q$ and 95% C.L. lower limits on the LQ mass as a function of $\beta = BR(LQ \to eq) $ are derived. The limits are 236, 205 and 145 GeV/c$^2$ for $\beta$ = 1, $\beta$ = 0.5 and $\beta$ = 0.1, respectively.