Pedestal and Common Noise
in the Test beam data Rudi
Ganz October 20th 1997
Here are some pictures generated from file #RUN00100.dat from
the 1997 test beam data. Here is for the Telescope number #1
ADC value versus channel number:
get the zip file with images HERE
Get projections here
 |
 |
| before the subtraction of pedestals* |
after subtraction |
( *generated as channel by channel average over the "pedestal tagged"
events)
The question now is whether there is common noise which we then have
to subtract on an event-by event basis. Common Noise means a correlated
change of the pedestal. So if one plots one the ADC value of one channel
against the other channel (in the same Read-Out chip) one can look for
such correlation's:
The histograms (as suggested by Russell) show the correlation
of the noise between each channel of one readout chip with all the
others. (So I take the first channel on the chip and plot it against channel
2-128; then I take second channel and plot it against 3-128 and so on ...)
Each telescope detector is read out by 5(4) 128 channel VA chips.
| Telescope #1 chip 1 |
Telescope
#1 chip 2 |
Telescope #1 chip 3 |
 |
 |
 |
| Telescope #1 chip 4 |
Telescope #1 chip 5 |
|
 |
 |
|
One can see no significant correlation in this telescope
detector -> no significant common mode noise. In fact when one tried to
subtract an event-by-event average on the non-hit channels one can improve
the already very narrow pedestal peak width (RMS. = 2.1 ch) to an RMS of
1.9ch. As expected from above, no significant improvement.