Home » Groups » STAR Protected

results of scanning the coincidence threshold

Updated on Wed, 2010-09-22 16:32. Originally created by rcorliss on 2010-09-22 16:32.

 pars out: 486.656769,0.023007,0.023105,1.796318

pars out: 480.609039,0.023020,0.022627,1.329064

pars out: 470.065521,0.022780,0.022259,1.013916

pars out: 454.951050,0.022581,0.021609,0.807482

pars out: 443.935516,0.022790,0.021248,0.594176

pars out: 428.203339,0.022248,0.020826,0.093737

pars out: 412.905334,0.022122,0.020529,0.048925

pars out: 399.082764,0.022029,0.020400,0.054971

pars out: 382.477509,0.022018,0.020195,0.060663

pars out: 362.948761,0.021567,0.020041,0.074229

pars out: 347.865143,0.021745,0.019983,0.075753

pars out: 331.398407,0.021664,0.019679,0.073868

pars out: 312.640808,0.021354,0.019339,0.094481

pars out: 294.956665,0.021068,0.019192,0.087120

pars out: 283.174408,0.021179,0.019471,0.077218

Figure 1, background rate from the gaussian fit to the vernier scan data for run 10097097.  The nominal threshold is the entry to the left of 0, and is clearly in the flat part of the curve.  The error bars are the uncertainties reported by minuit.
 
Since this background rate is believed to be non-flat, higher background rate should correspond to a less gaussian overall shape for the vernier scan, which in turn should make the uncertainties on the cross section larger, which is confirmed in figure 2.
 
Figure 2, fractional uncertainty in the BHT3coin cross section as a function of the ADC shift in the awayside threshold.  I fixed the offset, so that the nominal value now occurs right at zero.  The actual minimum is one bin to the left, but the difference is less than 1%.  At negative shifts the uncertainty comes from large nongaussian contributions in the background rate while at positive ADC shifts the upward shift is consistent with increasingly limited statistics.  The fractional uncertainty should be taken as a 'goodness of fit', relative rather than an absolute measure.
Groups: