CDC signal integration new
Cosmics, 50/50 gas, 2090V
Comparison with old (underpowered preamp) data is at CDC_signal_integration
Run numbers are 31809 (horiz) 31810 (45 deg) and 31811 (30 deg) at 3.0V, 2.09kV
dEdx attempts
time over threshold (signal duration)
dEdx without trackfitting
Ratio of fit MPV for horizontal : 30 degr : 45 degr data are 210 : 261 : 317 or 1 : 1.24 : 1.51 Ratio of vertical path lengths should be 1 : 1.15 : 1.41 or 1 : 1.16 : 1.45 after correcting for range of angles due to scintillator height and width.
Data below are for hit threshold 5 x pedestal sigma
Try to calculate dE/dx by dividing by track length (calculated in the plane parallel to the endplates)... (A) integrate signal for 700 ns (B) position of first max in amplitude (C) integrate signal up to position of first max in amplitude
Ratio of fit MPV for horizontal : 30 degr : 45 degr data are 16.9 : 21.4 : 25.7 or 1 : 1.27 : 1.52 Ratio of vertical path lengths should be 1 : 1.16 : 1.45
Ratio of fit MPV for horizontal : 30 degr : 45 degr data are 40.7 : 54.5 : 69.2 or 1 : 1.34 : 1.70
(Purple background section has histograms from previous section with the tilted data scaled by 1/cos(tilt angle))
Track length is calculated in the plane parallel to the endplates (no z information) so I scaled it by the tilt of the prototype for the 45 degree data, the upper cosmic trigger scintillator is high up enough that the cosmics are ~ vertical.
The ratio of tilted 45deg to horizontal integration means is 1.24 (old data) and 1.28 (new data); it should be 1 (as they are already scaled by track length) so this means the tilted integrations are 25-30% too large.
Landau fit MPVs are 16.9 (horizontal), 18.4 (30deg) and 18.1 (45 deg). The ratio of tilted 45deg to horizontal MPV of Landau fitted to amplitude/track length histogram is 0.99 (old data) and 1.07 (new data).
The ratio of tilted 45deg to horizontal MPV of Landau fitted to (integration to first maximum/tracklength) histogram is 1.07 (old data) and 1.18 (new data).
Look at time over threshold...
Search back from end of sample window until ADC value rises above a threshold, tried 3sigma, 5sigma, 8sigma (sigma= pedestal std dev). The initial hit selection is on 5sigma so all events will pass this.
Another way to find the peak length is to start at the end of the sample window and work towards the first sample, looking for a rise of 10sigma or more between 2 consecutive chunks of 5 samples summed together. Other similar approaches on this page finding signal duration