Difference between revisions of "Meeting-3-22-2018"
From GlueXWiki
(→Agenda) |
|||
| (17 intermediate revisions by 4 users not shown) | |||
| Line 1: | Line 1: | ||
| − | March | + | March 22, 2018 Drift Chamber meeting |
= Connection = | = Connection = | ||
| Line 12: | Line 12: | ||
# CDC settings for the rest of the run | # CDC settings for the rest of the run | ||
| − | ## Results from the CDC threshold scan run 41611 to 41614 (CDC hardware threshold 90, 110, 120, 140) | + | ## '''Results from the CDC threshold scan run 41611 to 41614 (CDC hardware threshold 90, 110, 120, 140)''' (Beni) |
### normalized rho yields at TH=120 increase by 4.3%, at TH=140 increase by 5.8%. Since these are different runs the data was normalized to the PS-flux | ### normalized rho yields at TH=120 increase by 4.3%, at TH=140 increase by 5.8%. Since these are different runs the data was normalized to the PS-flux | ||
### the rho yields seem to scale linearly with CDC hardware threshold. | ### the rho yields seem to scale linearly with CDC hardware threshold. | ||
| Line 19: | Line 19: | ||
### the software threshold of 140 leads to a rho yield increase of 3.7% while the hardware threshold of 140 to about 5.8%, this may be an indication that there is some shadowing effect. | ### the software threshold of 140 leads to a rho yield increase of 3.7% while the hardware threshold of 140 to about 5.8%, this may be an indication that there is some shadowing effect. | ||
### run 41614 with hardware threshold at 140, an increase of the software threshold from 150 to 160 does not increase the rho yield at a measurable level. | ### run 41614 with hardware threshold at 140, an increase of the software threshold from 150 to 160 does not increase the rho yield at a measurable level. | ||
| − | ### see logbook entry: https://logbooks.jlab.org/entry/3544993 | + | ### Note that a higher software threshold may lead to lower rho yields before kinematic fitting, however after kinematic fitting more rhos are left! This is particular significant when going from a software threshold of 110 to 140, see logbook entry: https://logbooks.jlab.org/entry/3544993 |
| − | ## Impact of the CDC Hit timing cut in the track reconstruction | + | ## '''Impact of the CDC Hit timing cut in the track reconstruction''' (Beni) |
### The increase in rho yield is significant. The cut is set at -60ns to 900ns | ### The increase in rho yield is significant. The cut is set at -60ns to 900ns | ||
### The higher the luminosity the higher the impact. At the highest luminosity (Run 41343) the impact was a factor of 2 increase in rho yield. | ### The higher the luminosity the higher the impact. At the highest luminosity (Run 41343) the impact was a factor of 2 increase in rho yield. | ||
### Similar effects are observed in the omega yields (both decay modes). | ### Similar effects are observed in the omega yields (both decay modes). | ||
| − | ## Impact of CDC Hit pruning, the removal of hits correlated in time with high amplitude hits that go into saturation on the same preamp card. | + | ## '''Impact of CDC Hit pruning, the removal of hits correlated in time with high amplitude hits that go into saturation on the same preamp card.''' (Beni) |
| − | ### A test on run 41343 with and without this pruning results in a increase rho yield of 4.6%. see also https://logbooks.jlab.org/entry/3543986 and https://logbooks.jlab.org/entry/3543986 | + | ### A test on run 41343 with and without this pruning results in a increase rho yield of 4.6% when pruning is applied. see also https://logbooks.jlab.org/entry/3543986 and https://logbooks.jlab.org/entry/3543986 |
| + | ## '''[https://halldweb.jlab.org/wiki/images/5/51/CDC_Efficience_vs_threshold.pdf Impact of threshold scan on CDC hit efficiency]''' (Alex) | ||
| + | ## '''[http://www.jlab.org/Hall-D/detector/fdc/spring2018run/sergey_CDC_2018_03_22.pdf Study of CDC raw signals]''' (Sergey) | ||
# Other | # Other | ||
| − | + | ## Track matched hit timing in MC: [https://halldweb.jlab.org/wiki/images/6/6c/Sdobbs_032218_sc_deltat_pt_proton_mc.png Proton in SC] [https://halldweb.jlab.org/wiki/images/e/e1/Sdobbs_032218_bcal_deltat_pt_proton_mc.png proton in BCAL] (Sean) | |
| − | + | ||
= Minutes = | = Minutes = | ||
| − | Participants: Naomi (CMU) Eugene, Beni | + | Participants: Naomi (CMU) Sean(FSU), Eugene, Beni, Simon, Alex A., Thomas, Sergey, and Lubomir (JLab). |
| − | = | + | = CDC settings for the rest of the run = |
| − | - | + | - Beni summarized the information related to the CDC thresholds from the hardware threshold scans and his software threshold scan (p.1.1 above). After long discussion, the conclusion was that the hardware threshold gives a little better yields for the rho production. At the same time there's indication for a plateau at ~140 still to be confirm with the hardware scan above 140. However, the main massage is that the effect is not so big compared to the effect of the timing (p.1.2). Filtering the noise (p.1.3.) also gives some improvements, but the main gain in the yields come from the timing. |
| − | - | + | - Alex A. showed that the effect of the hardware threshold on the hit efficiency is relatively small and acts mainly on the large drift distances (plot attached). |
| − | - | + | - Sergey (see presentation above) analyzed the raw signals and sorted them in several categories: normal, saturated, "photons", cross-talk. He studied their percentage for many raw-mode runs (unfortunately at different condition); for "standard" 2018 running the background events are about 35%. He estimated the shadowing effect to be about 3-5%. Sergey proposed different filtering methods (like Time-Over-Threshold) eventually to be implemented in the firmware. Simom is using Sergey's noise filtering on the raw data to see the effect on the rho production. |
| − | - | + | - Eugene asked about the amplitude distributions for hits at and out of the tracks. Naomi showed these distributions for different straws. She concluded that some of the straws (only several cards) are noisy and for these higher threshold (~140) would cut the noise, but that is not the case for the majority of the signals. |
| − | - | + | - The conclusion is that based on the information that we have, the hardware thresholds are not so critical (in contrast to the timing cut/window that is already implemented). We still want to have a minimal threshold of 110 plus some special thresholds for the noisy channels. Naomi/Beni will generate these thresholds for the rest of the run. This conclusion may not be valid for reactions other than rho production, or for different pattern recognition in the CDC. |
| − | + | = Other = | |
| − | - | + | - Sean studied in MC the BCAL vs Start Counter timings for proton tracks in the CDC. BCAL shows some deviations from the generated time at small p_T. |
Latest revision as of 11:04, 26 March 2018
March 22, 2018 Drift Chamber meeting
Contents
Connection
- Instructions for Bluejeans meeting connection
- Meeting ID: 290664653
Headline text
- To join via a Web Browser, go to the page [1] https://bluejeans.com/290664653.
Agenda
- CDC settings for the rest of the run
- Results from the CDC threshold scan run 41611 to 41614 (CDC hardware threshold 90, 110, 120, 140) (Beni)
- normalized rho yields at TH=120 increase by 4.3%, at TH=140 increase by 5.8%. Since these are different runs the data was normalized to the PS-flux
- the rho yields seem to scale linearly with CDC hardware threshold.
- CDC software threshold scan using run 41611 with hardware threshold at 90. First 3 files of the run are analyzed, same data so no normalization necessary. At 110 a rho yield increase of 3.3% is seen, at 140 a rho yield increase of 3.7% is seen, at 150 a rho yield increase of 3.4% is seen, at 160 a rho yield increase of 2.7% is seen.
- the rho yields do not scale linearly with software threshold! They saturate at about 140 (same as the highest hardware threshold)
- the software threshold of 140 leads to a rho yield increase of 3.7% while the hardware threshold of 140 to about 5.8%, this may be an indication that there is some shadowing effect.
- run 41614 with hardware threshold at 140, an increase of the software threshold from 150 to 160 does not increase the rho yield at a measurable level.
- Note that a higher software threshold may lead to lower rho yields before kinematic fitting, however after kinematic fitting more rhos are left! This is particular significant when going from a software threshold of 110 to 140, see logbook entry: https://logbooks.jlab.org/entry/3544993
- Impact of the CDC Hit timing cut in the track reconstruction (Beni)
- The increase in rho yield is significant. The cut is set at -60ns to 900ns
- The higher the luminosity the higher the impact. At the highest luminosity (Run 41343) the impact was a factor of 2 increase in rho yield.
- Similar effects are observed in the omega yields (both decay modes).
- Impact of CDC Hit pruning, the removal of hits correlated in time with high amplitude hits that go into saturation on the same preamp card. (Beni)
- A test on run 41343 with and without this pruning results in a increase rho yield of 4.6% when pruning is applied. see also https://logbooks.jlab.org/entry/3543986 and https://logbooks.jlab.org/entry/3543986
- Impact of threshold scan on CDC hit efficiency (Alex)
- Study of CDC raw signals (Sergey)
- Results from the CDC threshold scan run 41611 to 41614 (CDC hardware threshold 90, 110, 120, 140) (Beni)
- Other
- Track matched hit timing in MC: Proton in SC proton in BCAL (Sean)
Minutes
Participants: Naomi (CMU) Sean(FSU), Eugene, Beni, Simon, Alex A., Thomas, Sergey, and Lubomir (JLab).
CDC settings for the rest of the run
- Beni summarized the information related to the CDC thresholds from the hardware threshold scans and his software threshold scan (p.1.1 above). After long discussion, the conclusion was that the hardware threshold gives a little better yields for the rho production. At the same time there's indication for a plateau at ~140 still to be confirm with the hardware scan above 140. However, the main massage is that the effect is not so big compared to the effect of the timing (p.1.2). Filtering the noise (p.1.3.) also gives some improvements, but the main gain in the yields come from the timing.
- Alex A. showed that the effect of the hardware threshold on the hit efficiency is relatively small and acts mainly on the large drift distances (plot attached).
- Sergey (see presentation above) analyzed the raw signals and sorted them in several categories: normal, saturated, "photons", cross-talk. He studied their percentage for many raw-mode runs (unfortunately at different condition); for "standard" 2018 running the background events are about 35%. He estimated the shadowing effect to be about 3-5%. Sergey proposed different filtering methods (like Time-Over-Threshold) eventually to be implemented in the firmware. Simom is using Sergey's noise filtering on the raw data to see the effect on the rho production.
- Eugene asked about the amplitude distributions for hits at and out of the tracks. Naomi showed these distributions for different straws. She concluded that some of the straws (only several cards) are noisy and for these higher threshold (~140) would cut the noise, but that is not the case for the majority of the signals.
- The conclusion is that based on the information that we have, the hardware thresholds are not so critical (in contrast to the timing cut/window that is already implemented). We still want to have a minimal threshold of 110 plus some special thresholds for the noisy channels. Naomi/Beni will generate these thresholds for the rest of the run. This conclusion may not be valid for reactions other than rho production, or for different pattern recognition in the CDC.
Other
- Sean studied in MC the BCAL vs Start Counter timings for proton tracks in the CDC. BCAL shows some deviations from the generated time at small p_T.
