20190220-Analyzer-Mtg

From PREX Wiki
Jump to: navigation, search

Back to Main Page >> Analyzer_Meeting

previous meeting << >> following meeting

Logistic information

Meeting room at JLab: CC L210A (entrance is in the lobby across from the 2nd floor auditorium entrance)

 BlueJeans calling instructions:
 Toll-Free Number (U.S.&  Canada):  888-240-2560
 International toll number:         408-740-7256
 Bluejeans CODE:                  419 045 306
 Bluejeans link: https://bluejeans.com/419045306

Agenda

  1. Change meeting time?
    • CG: current time collides with the weekly beam scheduling meeting.
    • Cip will look back at the various timeslots we had considered and we'll
  2. "Operations" branch
    • Tao will extract the BPM pedestals from the run we'd gotten last night
  3. Pull requests
    • WD: He will try some further development for the TMapFile on the adaq machines; so far we see a simple test case is failing to properly link/load the ROOT libraries needed to run the TMapFile
    • PK: The fix to the analyzer to process the data stream from the 2ET system is in a pull request. Will pull that soon.
  4. online plots
    • Devi: He has some macros, and calling them from within panguin seems slow. Part of this is due to access through his laptop, but part is due to the macro call. Question is raised about whether calling a fully defined macro from within panguin is what really the best practice might be, or if its easier to just runt he macro directly.
    • KP: we should look at the online plots which PREX-1 had used, and start with those
  5. regression analysis
    • KP: When you do regression analysis, an hour is too long. We need to have a "mini-run" which is about five minutes at a time, and which can handle chunking and not leave orphan events if the last chunk is too small.
    • KP: How to associate "chunks" back to particular event ranges while doing second pass analysis.
    • KP: A goal would be to be able to produce the correlation matrix of all quantities and then to use that to determine slopes
    • KP: What is the interface for the correlator; Robert will prepare a summary of what we have for next meeting as a starting point
  6. feedback
    • KP: Can Amali prepare a summary of what the JAPAN feedback is doing, and what panffb is doing. I (Paul) will work with her about that.
  7. VQWK adcs
    • Ye Tian and Cameron will be evaluating the boards as we get them back from the electronics group
  8. Other issues
    • Run type classification
      • KP: basic types production, calibration, DAQ test. WAC should be evaluating between good, bad, and suspect. Probably usefule to have a comment on the type to flag BCM calibration, BPM calibration, etc. The goodness needs to be tagged for each arm independently.
      • KP: In addition to tuning the event cuts, we may need to also cut on event range. Maybe the mini-run data periods should be used to aggregate the data and the regression results. The mini-runs could then be tagged for goodness independent of each other. These would be similar to the runlets in Qweak.
    • What types of signals would we want to be able to set an alarm on (single defined variables, sums and differences of variables with scaling, arbitrary arithmetic functions of defined variables, ....)?
      • The alarm functionality will likely be based on the data handler/combiner classes. Do we need to be able to define ratios in the combiner class? KP: It might be nice, but we don't have a use case in mind right now, so let's defer action on this for now.
      • Do we need to be able to apply event cuts or alarms on combinations? Right now, event cuts are not defined on the combiner result variables. That ought to be simple to add.

Bob's DAQ tasklist

 TO-DO list for Parity DAQ
   updated Jan 4, 2019
Cameron & Bob: 1, 2, 4, 5, 6, 11, 12, 13
Tao & Cameron: 8, 9
Bob: 3, 6, 10
Cip, Cameron, Tao: 13  
1 Doing - Timing: mostly understood, still small mysteries (LNE)
2 Doing - Synchronization-check system. w/help from Ed Jastrzembski
3 Done - File naming, reset run num, finalize configs (INJ, CH, ALL) 
4 Done - Fix scalers on spectrometer crates.
5 To Do - Clean up cabling, labels, fasteners - First need to examine all signals
6 Clean up CRL -- eliminate so-called "dirty trick"
7 Check the 7200 scaler setup at injector, find a spare 7200.
8 Doing - Pedestal studies, aka cross-talk.  
9 Done - Test JAPAN. Need cable map & labels. Possibly maintain "pan" ?
10 Green Monster maintenance.
11 Doing - Test spares.  Repair bad modules.
12 Doing - Document everything.  Circuit diagrams, cable map, "how to".
13 Online software, a bit undefined since software is changing.
    (a) automatic production of run-dependent database
    (b) alarms like end-of-run voice (Bob)
    (c) feedback
    (d) coil-pulsing aka modulation aka dithering (Victoria)
    (e) panguin (GUI showing real-time plots)
    (f) prompt analysis
    (g) e-logbook entries 
    (h) archiving EPICs data
Accomplished
 * Revive all crates.
 * Deploy all ADCs needed and test each channel.
 * Build various DAQ configs with subsets of the DAQ crates.
 * Timing diagram, check timing of external signals and readout.
 * ET system upgrade, test rate limits, ensure deadtime zero.
 * Implement Bank structure.

Present

Paul K., Cip, Adam, Bob, Cameron, Jinglong, Kent, Victoria, Robert, Sakib, Ye Tian, Devi, Tao, Wouter

Excused