Sphenix-calibration-l mailing list

[Hugo Pereira Da Costa <hugo.pereira-da-costa AT cea.fr>]

Hi all,

The notes from the last meeting are available at https://indico.bnl.gov/event/10802/attachments/32758/52300/notes.txt and below

best,

Ross and Hugo

-----------------------------------------------------------------

1/ Reconstructed Central Membrane Clusters (Ross)

Ross showed CM stripes created as a separate g4hit container, passed through electron drift code and reconstructed from output clusters.  
Slide 3: 
C: crash with edrift->Detector("TPC") should not happen (cannot reproduce). Need further investigation as it might hide deeper problem with local compilation (e.g. inconsistent headers)
Q: how are G4Hits identified ? 
A: starting from an arbitrarily large fixed ID
C: could create a problem, conflict with existing G4Hits from e.g. GEANT. However G4Hits from CM are stored in different map. 
Conclusion: no conflict possible, and no need to start ID with an offset. could start at zero
If it is decided to put it in the main TPC G4Hit map, ids should start after the last one from current event.

Slide 4:
Additional Q: what type of events is thrown, how does the event deposited charge compare to that of the CM ? 
Additional C: with this in place it should be easy, at the G4Hit level to compare the charge generated by the CM events to that from MB + PileUp, assuming a given trigger strategy (once per event or less) to quantify possible impact on SC distortions.

Slide 5: 
Q: does this include distortions ? 
A: yes

Slides 6 to 10:
Some Clusters are lost during the cluster-pairing.  Work in progress.
Q: can one loosen the clustering cuts to recover missing clusters ? 
A: clustering cannot be loosened -- the thresholds are hardware issues, so there's no way to easily 'allow this threshold to be lower if the adjacent row has a clear cluster'. 

C: some CM stripes produce only one cluster, which we would expect if there is a 1.5*pad row distortion which aligns the stripes that usually straddle two pad rows.  With a little focusing, or with losses due to pedestals, the neighbor doesn't appear.

Q: should we do something with the region of the acceptance where only one raw is hit by the CM clusters ? 
A: they correspond to regions of the acceptance where the distortions are about 1.5cm, so that CM and readout pads align perfectly. 

Tom and Ross consider what can be done to actually resolve the stripe centers -- can beat 1/sqrt(12) because we know the true shape, but needs modeling.  Can also lengthen or shift the pads if needed to avoid the beat pattern. Shifting the pads would strongly rely on the static distortions from simulations. How close those are to the one we will see in data is unclear. For instance there can be further alignment problem that changes the static distortion, and move the position at which the CM matches the rows, nullifying the effort.

Lengthening the CM pads would make the centroid reconstruction more complicated, and would create cases where you get clusters of size 3 along r, making the algorithm more complicated.

Additional Q: should there also be some clustering in place in the phi direction (or is there already) ? What is the expected phi cluster size for a given CM pad ? How does that change with phi distortions in place ?

(long) Discussion on trigger strategy for the CM events:
cannot have the CM flash exist at eta=0 with respect to a real event or it will swamp the real eta=0 tracks.
Tom suggests tying in the firing to 5* sigma_z so that it is never near physics tracks. (or could use x*max electron drift time ?) 
Discussed whether we need a trigger busy after the laser flash for some window to keep the laser from interfering with subsequent events.  
Tony points out it's easier to throw those events away if they're trouble rather than not taking them.  

Additional comment: Note that the nuisance goes both ways: having track hits in the midst of CM hits will also affect the CM reconstruction, and possibly locally deteriorate the clustering, and ability to reconstruct the distortions.

Christof notes that it is hard to calculate the amount of luminosity impacted by either choice.  Needs a toy model.
Ross says that original interest in having the steady beat of laser flashes (for time series work and for uniformity of space charge) no longer seems as necessary now that we understand how the various components combine

Discussion on how to deal with the CM hits in the "standard" track finding.
Biggest impact will be on track seeding.
Ross hopes there is some way to work around the flash-pancake without killing tracks that pass through it. At the z "slide" occupied by the CM hits, there is probably no way that a seeder can disentangle track hits from CM hits. Easiest is probably to remove all the hits in such pancakes from the seeding altogether ...

Misc: 
Need to add code to allow the pancake to be offset in time (not z).  (R: For play purposes, probably suffices to move the z position of the initial hits, though in reality we would want to diffuse them fully from the CM, then just shift the resulting z.)


2/ Biases in distortion reconstruction (Hugo): postponed

On 2/8/21 9:41 AM, Hugo Pereira Da Costa via sPHENIX-calibration-l wrote:

    Dear All,
      
      
      The next meeting on outer-tracker studies and spacecharge
      distortions in the TPC will take place tomorrow, Tuesday, Feb. 9,
      at 11AM EDT
      
      
      The indico page is at https://indico.bnl.gov/event/10802/ and
      connection details are below.
      
      
      Contributions are welcome !
      
      
      Best,
      
      
      Hugo
      
      
      ------------------------------------------------
      
      
      Meeting URL
      
      https://bluejeans.com/676999731?src="join_info
      
      
      Meeting ID
      
      676 999 731
      
      
      Want to dial in from a phone?
      
      
      Dial one of the following numbers:
      
      +1.408.419.1715 (United States(San Jose))
      
      +1.408.915.6290 (United States(San Jose))
      
      (see all numbers - https://www.bluejeans.com/premium-numbers)
      
      
      _______________________________________________
      
      sPHENIX-calibration-l mailing list
      
      sPHENIX-calibration-l AT lists.bnl.gov
      
      https://lists.bnl.gov/mailman/listinfo/sphenix-calibration-l