Recoil Detector FAQ
Page maintainer: Andreas
 |
This is a working page which will continuously change. The content is not reviewed and is intended for people working on this specific topic. |
Back to Main Recoil Page
Here is a collection of frequently asked questions related to the
Recoil Detector. Most of what can be found here are condensed versions
of discussions on the
recoil-commissioning
mailing list. The idea is that once a discussion on the Recoil
Detector has come to a conclussion, a summary should be added
to this page.
What do abbreviations like SFT mean?
Several abbreviations are used for the Recoil Detector components.
RD - Recoil Detector
SSD, Silicon - Silicon Strip Detector
SFT, SciFi - Scintillating Fiber Tracker
PD - Photon Detector
How are dgDETS numbers assigned to Recoil clusters and spacepoints ?
For the SSD we have 16 sensors with p- and n-side signals, p- and n-side clusters in one sensor have the same dgDETS and are distinguished by strip numbers (1 to 128 for the p-side, 1001 to 1128 for the n-side), so we expect 16 dgDETS for clusters and 16 dgDETS for spacepoints.
For the SFT in the inner layer we call "spacepoint" a combination of at least one cluster in the parallel layer and at least one cluster in the stereo layer. The dgDETS for spacepoint is set to dgDETS of the innermost parallel layer. Therefore, we can have SFI1 in case of signals in the first or both SFT parallel layers of the inner barrel, SFI2 in case of no signal in SFI1 and SFO1 in the outer barrel. So, three in total.
For the photon detector even one cluster can form a spacepoint and dgDETS is set according to the innermost cluster. For example, if we have a spacepoint from clusters in the B and C layers, dgDETS for the spacepoint is set to dgDETS of the B layer. So, we also have three possible dgDETS for spacepoints in total.
Please be aware that the dgDETS numbers are different in data and MC. You may find a list here (script suited). Using this list of hardcoded dgDETS in an analysis is a potential source of errors and therefore not recommended. Please access the geometry server (DATA) or geometry file (MC) to get the correct numbers instead.
What is the unit of the energy value in a rdCluster and a rdSpaceoint ?
For the SSD the spacepoint energy is either the mean of the p- and n-side energy values of the used clusters (the usual case) or the p-side or n-side energy deposit alone (this is the case if one of the sides does not allow a good energy determination). For the clusters and spacepoints the unit is GeV.
For the SciFi and PD the energy of a spacepoint is simply the sum of the energy of clusters used to form the spacepoint. The unit is GeV as well.
How is the internal alignment of the Recoil Detector components done ?
First, the information from the dedicated Scintillating Fiber Tracker (SFT) alignment run at DESY 22 test beam (see [1] for details) was implemented to xtc. As a results for all combinations of SFT parallel and stereo layers phi angle and z coordinate is read out from a text file. Alignment has been checked on cosmic data.
Second, each sensor of the Silicon Strip Detector (SSD) was aligned relative to the SFT using straight tracks from dedicated runs with the Recoil magnet switched off ("magnet-off" runs). Photon Detector (PD) was also aligned relative to the SFT using straight tracks from magnet-off runs.
What do iMethod numbers in rdTrack table mean?
See a short description here [2]
How is the Photon Detector used in the Recoil tracking?
All tracking methods use SSD and SFT information. In the Photon Detector only clusters and spacepoints are reconstructed. Coordinates of such clusters (spacepoints) are compared with coordinates predicted by tracks reconstructed in the SSD and SFT. If clusters (or spacepoints) are close enough to tracks these clusters (spacepoints) are related to corresponding tracks. If not, then such clusters (or spacepoints) are called "trackless" and have no relations to tracks. For example, photons from pi0 decay should produce "trackless" clusters.
Which productions with Recoil data are available?
Recoil data have been taken in the years 2006 and 2007. Look here for an overview which information is contained in which production and some data quality plots.
How many data (statistics) were taken in total with the Recoil detector?
Preliminary Statistics (Status: Jan.2008)
What are dimensions of the Photon Detector?
The information about dimensions was sent to the recoil-commissioning list by Charlotte [3].
How is the beam position taken into account in the Recoil tracking?
Beam position is used in the track fitting (iMethod 7, 701). The beam position calculated using dedicated Recoil beam finder in the Recoil coordinate system is compared with the beam position determined with the beam position monitors (BPMs) installed at HERMES. From this comparison connections between the Recoil coordinate system and the BPM coordinate system is determined and in productions the informations from BPMs which is available for each run is used.
What materials are used for the SFT?
The latest information can be found here [4].
What is the location of SSD sensors in the HERMES coordinate system?
Azimuthal (phi) coordinates of the SSD sensors centers are the following (without misalignment):
7*PI/4 - First quadrant of the SSD (sensors S1I1, S1O1, S2I1, S2O1)
PI/4 - Second quadrant of the SSD (sensors S1I2, S1O2, S2I2, S2O2)
3*PI/4 - Third quadrant of the SSD (sensors S1I3, S1O3, S2I3, S2O3)
5*PI/4 - Forth quadrant of the SSD (sensors S1I4, S1O4, S2I4, S2O4)
What is the difference between Proton and Pion hypothesis in rdTrack table?
Pion and Proton hypothesis has nothing to do with particle identification (PID). Protons and pions have different energy losses in the Recoil detector materials, therefore, the curvature of tracks with the same momenta is different for pions and protons. It is taken into account during the track fitting using momentum- and angular- dependent corrections of phi coordinates measured in XY plane. As a result, momentum and phi angle reconstruction is slightly different for pions and protons. The difference becomes small with increase of momentum. When PID is done using energy deposits in the SSD, SFT and PD one of the track parameters in rdTrackPar table will be used for final analysis.
Which are the dead channels in the SFT?
- electron beam data: 244 channels. SF01: fibers 857-977; SF03: fibers 849-968
- positron beam data: 367 channels. SF01: fibers 795-977; SF03: fibers 788-968
- Weilin's TRM presentation 14.5.2008
Which geometry file should I use when I analyze recoil data?
- Beni's mail 7.5.2008
Why does my uDST code crash when I read 2006/2007 data?
In case your uDST code reports something like this:
ZEBRA internal error Program will STOP because of catastrophic error
do the following...
As the tables related to the Recoil detector are large and take up a lot of space, one has to increase the memory available to the ADAMO TAble Package (TAP). To do this check your makeddl script and search for a line that starts with
for/set initap INITAP:XYZ
where XYZ is the memory size used. Increasing the value to 30000000 should cure the problem.
How can I convert phi angle to fiber number for the SFT?
The conversion from fiber numbers to phi angles is done in the sft_spacepoint_calc routine in xtc (sft_spacepoints.c). If alignment information is included, phi also depends on the z coordinate.