Status of L1 and L2 analyses for the FPS trigger: (A. Lucotte)
=================================================

(1) A d0-note 3398 describing the result from the Beam Test 
    has been released. It constitutes the first test of the 
    simulations used to estimate selection/rejection rates in
    the trigger studies and provides an absolute calibration of 
    the detector as well. 

(2) Answer to an adressed question:  
    The FPS readout scheme is as follows: Photons are carried from
    the scintillator fiber to the VLPC where they are converted 
    into electrical signal. At that point, the signal is charge 
    divided into two arms with a ratio typically 1:10. In each 
    of the amrs, the signal is shipped to its own set of SIFT 
    chips and SVX chips. The SIFT chips are used to provide 
    a fast L1 trigger via the setup of discriminator ranges.   
    The high gain arm allows us to detect a MIP while the low gain 
    arm signals higher energy depositions, like those in a E.M. shower.  
    The L1 algorithm is based not only upon the detection of a MIP 
    (in front of the lead) and a shower threshold (behind the converter)
    but also on a isolation threshold.  
    The question from electronic group (M. Johnson, Fred Borcherding..)  
    was whether it was possible to define a isolation threshold 
    in a ratio approx. 1:10 with respect to the high threshold used 
    for the shower. L1 studies show that a redefinition of the isolation 
    window (strip index beside the central strip of the shower) results 
    in a ratio 1:7 for this threshold, while keeping a reasonnable 
    e- selection efficiency and a ratio e-/pi0 of 3. 

(3) Status of L2 trigger (M. Bhattacharjee, A. Lucotte)
    A study regarding a L2 trigger using ANALOG information has 
    been undertaken, assuming that the full information is used 
    (all strips readout). The goal is to use the energy dependence 
    of the e-/pi0 shower shape to set adjustable thresholds in order 
    to get a good rejection in the whole energy range. The main 
    outcomes of this preliminary work are: 
    i - The correlation between deposited energy and true energy 
        of the e- shows that a L2 energy-dependent shape study 
        may improve the rejection capability for low energetic 
        particule only (<30 GeV or ~<10 GeV in Pt).
    ii- M. Bhattacharjee shows that rather than the total 
        energy deposited in the shower, the cluster energy formed
        by the sum of the 3 central strips provides a more important 
        discriminating power. (less sensitive to fluctuations of the 
        tails of the distributions)      
    iii-An isolation window has to be adjustable as function of the 
        true energy of the particle, as well as the isolation threshold 
    iv -These studies will lead to the creation of a lookup table 
        providing the 90% probability (1=90% e- prob., 0=no ele.) that  
        the signal is an electron, given the values of E(cluster) and E(iso). 
        No lookup table has been developped yet. 

    Preliminary results confirm that an improvement in the lower Pt range 
    e- efficiency is reachable (20%) at L2 with respect to L1. 
    The ratio e-/pi0 does not gain much though. 
    
    What remains to be done concerns the lookup-table and the use 
    of realistic events (single particle + 2 mbias) to set the thresholds.