Re: Centrality (was: ZDC cut efficency vs multiplicity)

From: Christian Holm Christensen (cholm@hehi03.nbi.dk)
Date: Tue Jan 09 2001 - 09:47:52 EST

  • Next message: Flemming Videbaek: "more on dn?deta using spectrometer"

    Hi Bjorn et al, 
    
    On Tue, 9 Jan 2001 09:18:10 -0500 (EST)
    Bjorn H Samset <bjornhs@rcf.rhic.bnl.gov> wrote
    concerning ": Re: Centrality (was: ZDC cut efficency vs multiplicity)":
    > I have tried to impose a vertex cut of +-30 cm on the ZDC/Mult plots, and
    > to my eye it makes no difference to the spectra in this case. 
    
    You should also be aware of the asymmetry of the tile layout, as
    reported by J. H.. This makes the ADC/Vz distribution skewed to the
    left. Hence, a symmetric, at least one as wide as +/- 30cm will be
    errouneous. 
    
    > Maybe the cut should be narrower - this is being tested now. The
    > plot I gave you is the one without a cut, since it had better
    > statistics 
    
    Low statistics is obvious the problem of all this. If, as Flemming
    claims trigger 4 is really the minimum bias trigger, then one should
    use inclusive trigger 4 (as in the upper 8 bits of the trigger word)
    events to determine the centrality thresholds. However, for later
    runs, where the downscale factor of trigger 4 is very high ~ 100, this 
    will pose the problem of low statistics. 
    
    However, one can do the analysis on runs with a lower scale factor,
    and compare the threshold values with those from later runs, and my
    guess is that we'll find the thresholds at similar ADC values. 
    
    > - for one with cut have a look at
    > www.fys.uio.no/~bjornhs/Run2235_25k_vtxcut.ps 
    > This one has 25k events. The other difference is that here the top left
    > plot is trigger 4 mult and trigger 6 mult superimposed. As Claus pointed
    > out to me, we seem to have events with a high mult but no trigger 6! Any
    > ideas on this?
    
    No, I'm afraid not. We've seen something similar and are at least as
    puzzled as you. If you plot the histogram with errors, you'd be even
    more dazzled! If the two distributions were to overlap within errors,
    I'd feel comfortable in saying that is some statitical fluc, but since
    they don't, I'm puzzled - I said that already - hell I'll say  it
    again: I'm puzzled!
    
    One reason could be, that trigger 6 is inefficient -
    that is, in some events the trigger just wasn't set, even though it
    should have, due to some error, either in the DAQ or the Beam-Beam
    counters themselves. Another possibility, would be that since the
    Beam-Beam arrays, in particular the big tubes, only cover a small
    fraction of the solid angle, and may not be hit in some event, either
    due to some interresting physics like fluctuations, or just because
    the emittion from the collision is stocastic - i.e., statistical
    fluctuations. But hey, I'm only speculating.  
    
    Yours, 
    
    Christian  -----------------------------------------------------------
    Holm Christensen                             Phone:  (+45) 35 35 96 91 
      Sankt Hansgade 23, 1. th.                  Office: (+45) 353  25 305 
      DK-2200 Copenhagen N                       Web:    www.nbi.dk/~cholm    
      Denmark                                    Email:       cholm@nbi.dk
    



    This archive was generated by hypermail 2b29 : Tue Jan 09 2001 - 09:48:24 EST