Hits 2001

Looking at 82 events from st_physics_2192029_raw_0004.daq. Online ASIC thresholds (3,2,5,0) (vallo, nlo,valhi,nhi), pedestal offeset 10. The files was off. Off line Common mode noise subtraction done via subtracting the mean ADC value of a timebucket if >50 andoes fired. Offline ASIC threshods were (2,1,5,0) (vallo, nlo,valhi,nhi) and the Inverse product cut 8. A hit is deemed good if flag <4

Fig 1
Clearly Hybrids 316 and 317 (Barrel 3 ladder 3 Wafer 6 Hybrids 1 and 2)are noisy and I've removed them from the rest of these plots.

Fig 2
Fig 2. shows the number of hits in the whole detector vs event number. It can be seen that there are some evnts with a much higher hit rate than the rest. Fig. 3 shows a plot of the correlation between the TPC hits and good SVT hits. You can see there is a weak correlation. Investigation is still being made to see if a stronger correlation can be produced. Some possible avenues are that currently all vertices are allowed and the vertex distribution for this data is quite wide. Large z vertices still leave many hits in the TPC but much fewer hits in the SVT, hence the correlation is reduced. Also the wrong T0 is currently used for hte SVT good hit evaluation ( we assume 0 here and it is belived to be closer to 6) this will allow bad hits in and kill good hits a longer drift. A better correlation to look at would be tracks from the TPC believed to pass through the SVT and hits in the SVT.
Fig 4

Fig 4
Here you can see the expected drop in peak ADC vs drift time. The profile histogram ignores the lowest few bins as I think these are just noise. It is good to see that even at long drift the hits don't disappear into the noise, even though the profile histogram is not wonderfully representative of the data even ignoring the noise at low ADC values. We could therefore set our high threshold slightly higher and not loose signal (except for the ripple argument see here

Fig 5
Clearly there is are some noisy hybrids here which is firing everything and causing one large hit to be found around timebucket 60. The bad hybrids are identified as hybirds 19,151,165,269 and 397. The rest of the plots are drawn without these hybrids.
Helen Caines
Last modified: Mon Jul 23 18:54:06 EDT 2001