• Main INDEX
  • Monthly INDEX
  • PREV
  • NEXT
    Make New Entry, Make Followup Entry

    User name brads

    Log entry time 15:30:21 on February 16, 2009

    Entry number 261511

    This entry is a followup to: 261414


    keyword=re: Analysis of Bigbite Gas Cherenkov

    1. The shift in coincidence cut 2 is probably because the signal is a little bit out of the ADC gate. I guess the best way to look at it is to select good events in the adjacent mirror to see this effect. It is not convincing to me that they are the events hitting other mirrors.
    If the track is near the edge of a mirror then the photons in the Cerenkov cone get split across multiple mirrors. This could be checked by histogramming the analog group sum of the adjacent mirrors and seeing if the shift goes away.
    2. The FASTBUS 1881 ADC resolution is 50 fC/channel. The VME 791 is 0.1pC /channel. I guess the number of channel should be divided by 2 for the new ADC, right? (Brad may comment on this) Thus, the 1 photon peak should be 60?
    I'm not sure which run is analyzed here. We switched to 1881 ADCs at 9pm on Feb. 13. So runs >1400 have 50 pC/chan, runs before use the v792 at 100 pC/chan.
    3. One should plot DBB.BBcerT06 vs DBB.t2. In this case, a clear correlation can be found. Since, the data are mostly from T2. Most likely the events in the coincidence timing 2 region is random coincidence (but give the trigger, since the T6 generate the gate and cerenkov signal define the timing). The first peak should be coincidence, I guess.
    There's no AND required for the DBB.BBcerTXX signals so you don't get the T1+T3 type randoms box with a peak in the middle (all randoms are registered by the TDC with equal probability(*)). The peak in region 1 is when that mirror has the same timing as the L1A. See Figure 1 (this message), where the T2 timing walks along with the BBcerT06 timing vs. the T6 timed L1A. The horizontal striation features between y-axis channels 610:620 is weird. Maybe adjacent Cerenkov mirrors carry the T2 timing there? For bonus points, can someone explain why T6 should walk vs the Cerenkov timing, but NOT walk vs. the MWDC signals. Does the MWDC analysis software already have offsets applied that take out the row-to-row relative timing shifts present in the shower? (*) With the usual caveats associated with electronic DT issues.

    A copy of this log entry has been emailed to: yqiang@jlab.org,brads@jlab.org,meziani@jlab.org,xqian

    Figure 1