Difference between revisions of "CDC Pedestal Studies"
(→Pedestal Subtraction) |
|||
(13 intermediate revisions by the same user not shown) | |||
Line 1: | Line 1: | ||
=== Pedestal Subtraction === | === Pedestal Subtraction === | ||
− | + | In this study the file hd_rawdata_003221_000.evio from run 3221 is used. It is a BCAL trigger run with the collimator in blocking position so no beam but only cosmics. Events are selected with the following signature in the CDC looking for hits with no signal in the first 125 samples. | |
* [[File:wf.gif|400px|tumb|| Wave form of CDC signal]] | * [[File:wf.gif|400px|tumb|| Wave form of CDC signal]] | ||
Line 7: | Line 7: | ||
<math>D = I - P_{i}/N_{i}*80.</math> | <math>D = I - P_{i}/N_{i}*80.</math> | ||
where <math>N_{i}</math> is the number of samples used to calculate the pedestal and 80 is the number of samples in the Integral I. | where <math>N_{i}</math> is the number of samples used to calculate the pedestal and 80 is the number of samples in the Integral I. | ||
− | In the case of no signal the the distribution D is expected to be around zero with a sigma that reflects the resolution and depends on the number | + | In the case of no signal in the region I the distribution D is expected to be around zero with a sigma that reflects the resolution and depends on the number samples <math>N_{i}</math> used in determining the pedestal. |
− | samples <math>N_{i}</math> used in determining the pedestal. | + | As an example the distribution, with 4 samples and 32 samples in the pedestal calculation, looks like this: |
− | As an example the distribution with 4 samples in the pedestal calculation looks like this: | + | |
* [[File:ped4sub.gif|400px|tumb|| Pedestal subtraction distribution]] [[File:ped32sub.gif|400px|tumb|| Pedestal subtraction distribution]] | * [[File:ped4sub.gif|400px|tumb|| Pedestal subtraction distribution]] [[File:ped32sub.gif|400px|tumb|| Pedestal subtraction distribution]] | ||
Line 27: | Line 26: | ||
| 64 || 40.9 || 680 || 84.6 || 685 | | 64 || 40.9 || 680 || 84.6 || 685 | ||
|} | |} | ||
+ | |||
+ | Note this is not what one would expect. The mean should be around zero! | ||
+ | |||
Looking just at the pedestals you get the expected results like here, a pedestal determined over the first 16 samples: | Looking just at the pedestals you get the expected results like here, a pedestal determined over the first 16 samples: | ||
− | * [[File:pedestal0.gif|400px|tumb|| Pedestal over | + | * [[File:pedestal0.gif|400px|tumb|| Pedestal over 0-16 samples]] [[File:pedestal6.gif|400px|tumb|| Pedestal over 96-112 samples]] |
+ | |||
+ | However if you look at the distribution of the above two pedestal calculations P6-P0 you get the following: | ||
+ | * [[File:peddiff.gif|400px|tumb|| Pedestal difference P6-P0]] | ||
+ | |||
+ | The following table shows the mean and sigma of the distributions when P0 is subtracted from all the other | ||
+ | pedestal distributions, once in full floating point and once with integers only. | ||
+ | {| class="wikitable" | ||
+ | |- | ||
+ | ! Diff !! mean F !! sigma F !! mean I !! sigma I | ||
+ | |- | ||
+ | | P1-P0 || 2.5 || 201.5 || 2.8 || 199.7 | ||
+ | |- | ||
+ | | P2-P0 || 26.6 || 254.9 || 27.7 || 257.9 | ||
+ | |- | ||
+ | | P3-P0 || 23.1 || 272.3 || 24.9 || 274.1 | ||
+ | |- | ||
+ | | P4-P0 || 25.2 || 280.0 || 25.9 || 275.0 | ||
+ | |- | ||
+ | | P5-P0 || 30.3 || 277.5 || 28.8 || 276.8 | ||
+ | |- | ||
+ | | P6-P0 || 21.3 || 280.1 || 19.7 || 282.1 | ||
+ | |} |
Latest revision as of 09:02, 24 June 2015
Pedestal Subtraction
In this study the file hd_rawdata_003221_000.evio from run 3221 is used. It is a BCAL trigger run with the collimator in blocking position so no beam but only cosmics. Events are selected with the following signature in the CDC looking for hits with no signal in the first 125 samples.
The first test was done by using the region (blue) to determine a pedestal and the region I (red) to determine an Integral and then subtract the proper pedestal from the Intergral as follows: where is the number of samples used to calculate the pedestal and 80 is the number of samples in the Integral I. In the case of no signal in the region I the distribution D is expected to be around zero with a sigma that reflects the resolution and depends on the number samples used in determining the pedestal. As an example the distribution, with 4 samples and 32 samples in the pedestal calculation, looks like this:
mean F | sigma F | mean I | sigma I | |
---|---|---|---|---|
4 | 102.2 | 1293 | 96.5 | 1276 |
16 | 119.0 | 1121 | 163.4 | 1120 |
20 | 121.1 | 1080 | 145.1 | 1080 |
32 | 102.4 | 984 | 159.0 | 978 |
64 | 40.9 | 680 | 84.6 | 685 |
Note this is not what one would expect. The mean should be around zero!
Looking just at the pedestals you get the expected results like here, a pedestal determined over the first 16 samples:
However if you look at the distribution of the above two pedestal calculations P6-P0 you get the following:
The following table shows the mean and sigma of the distributions when P0 is subtracted from all the other pedestal distributions, once in full floating point and once with integers only.
Diff | mean F | sigma F | mean I | sigma I |
---|---|---|---|---|
P1-P0 | 2.5 | 201.5 | 2.8 | 199.7 |
P2-P0 | 26.6 | 254.9 | 27.7 | 257.9 |
P3-P0 | 23.1 | 272.3 | 24.9 | 274.1 |
P4-P0 | 25.2 | 280.0 | 25.9 | 275.0 |
P5-P0 | 30.3 | 277.5 | 28.8 | 276.8 |
P6-P0 | 21.3 | 280.1 | 19.7 | 282.1 |