Timing adjustment for Tower crates
  • Example of tower ped subtracted ADC spectra for crate timed in and the one out of time.
    Conditions for the run 5009138 are described in eLog #523. Tower setup: ( RHIC delay 80ns & FIFO=19) x 6 crates. Trigger: MinB. Total of ~600 events in each plot. xRay() only assures crate ID matches its place in the data stream and the token Run 5009138 condition are described in eLog #523. Towers in crates 1-4 have HV, 5 & 6 were without HV. Tower setup: ( RHIC delay 80ns & FIFO=19) x 6 crates. Trigger: MinB. Total of ~600 events in each plot. xRay() only assures crate ID matches its place in the data stream and the token is right.
    Pedestals are subtracted.

    Single tower ped subtracted spectra for crate 4 with good timing.


    Single tower ped subtracted spectra for crate 1 with the timing off by 80ns.

  • Method - integrate ADC spectra over range [10,100], ped subtracted. To measure of how well given crate is timed in the ratio of integral to the total number of used events vs. delay is plotted. The last year formula is used
    delay=delay + 106*(FIFO-19).
    Most of data are taken with FIFO of 19. Data taken with FIFO of 20 do not match the pattern. Any feedback is welcome. I'm not sure where to put data with delay of 110 --> at 110 ns or at 4 ns ?
    For each crate we selected 10 towers in the eta bin #7 (sub-sectors DEABCDEABC for respective crates) to cover a reasonable wide range of channels in a crate.

    CRATE 1 , best delay 10 ns, FIFO=19


    CRATE 2, best delay 0 ns, FIFO=19


    CRATE 3, best delay 70 ns, FIFO=19


    CRATE 4, best delay 70 ns, FIFO=19


    CRATE 5, best delay 40 ns, FIFO=19 , some towers did not responded


    CRATE 6, best delay 30 ns, FIFO=19
    Macro used plTiming.C