Impact of BC2 and BC3 on simulated data, comparison SAMPA MPW2 design review II Meeting Konstantin Münning Universität Bonn Helmholtz-Institut für Strahlen- und Kernphysik 01.04.2015
BC2, BC3 operation principle BC2 operation BC3 operation amplitude [ADC] amplitude [ADC] 160 160 140 140 120 120 100 100 80 80 60 60 510 520 530 540 550 560 570 510 520 530 540 550 560 570 samples samples • BC2 is calculating average from values within thresholds • BC3 is following the signal limited by slopes 2
BC2, BC3 operation principle detail BC2 operation detail BC3 operation detail 85 85 amplitude [ADC] amplitude [ADC] 80 80 75 75 70 70 65 65 60 60 55 55 540 542 544 546 548 550 552 554 556 558 560 540 542 544 546 548 550 552 554 556 558 560 samples samples • BC2 stays constant during peak data • BC3 constantly follows but in a simple deterministic way 3
simulation parameter signal peak amplitudes CM peak amplitudes 3 10 × 250 count count mean=500 mean=150 sigma=50 sigma=100 600 200 500 150 400 300 100 200 50 100 0 0 0 200 400 600 800 1000 0 200 400 600 800 1000 amplitude amplitude • sampling frequency=10MHz, t p =160ns • charge rate=2GHz (constant) • Gauss and Landau distributed amplitudes 4
minimal PSA signal peak amplitudes • simple minimal PSA used for count mean=499.65±0.51 peak evaluation sigma=50.58±0.37 600 mean=505.09±0.66 • peak amplitude computed sigma=54.64±0.49 500 by peak area 400 • signal spectrum is altered 300 by minimal PSA • no special handling of piled- 200 up events 100 • Gauss fjt of histogram 0 0 200 400 600 800 1000 amplitude • signal spectrum of PSA is used as a reference to evaluate the impact of the fjlters to the signal • error values from fjt 5
without common mode efgect (1) signal peak amplitudes • results of both fjlters similar count mean=505.09±0.66 sigma=54.64±0.49 • moving average fjlter BC2 is 600 mean=504.47±0.66 slightly closer to ideal value sigma=54.58±0.49 500 mean=499.75±0.66 • slope based fjlter BC3 sigma=54.98±0.49 400 without slope correction 300 shows larger deviation 200 100 0 0 200 400 600 800 1000 amplitude 6
without common mode efgect (2) signal peak amplitudes • slope based fjlter BC3 with count mean=505.09±0.66 simple slope correction is sigma=54.64±0.49 600 mean=504.47±0.66 signifjcantly closer to ideal sigma=54.58±0.49 500 value mean=502.40±0.66 sigma=54.52±0.49 400 • better correction is possible, not needed for SAMPA 300 submission 200 100 0 0 200 400 600 800 1000 amplitude 7
with common mode efgect (1) signal peak amplitudes signal peak amplitudes count count mean=499.65±0.51 mean=499.65±0.51 sigma=50.58±0.37 sigma=50.58±0.37 600 600 mean=505.09±0.66 mean=405.75±0.69 sigma=54.64±0.49 sigma=58.61±0.50 500 500 400 400 300 300 200 200 100 100 0 0 0 200 400 600 800 1000 0 200 400 600 800 1000 amplitude amplitude • without baseline correction there is a signifjcant deviation of mean value when common mode efgect is present (right) compared to values without common mode efgect (left) 8
with common mode efgect (2) signal peak amplitudes signal peak amplitudes count count mean=505.09±0.66 sigma=54.64±0.49 600 600 mean=504.80±0.66 sigma=54.72±0.50 500 500 mean=503.293±0.66 sigma=54.68±0.49 400 400 300 300 200 200 100 100 0 0 200 400 600 800 1000 420 440 460 480 500 520 540 560 580 600 amplitude amplitude • with baseline correction the mean value is restored similarly well by both fjlters 9
with common mode efgect (3) signal peak amplitudes signal peak amplitudes count count mean=505.09±0.66 sigma=54.64±0.49 600 600 mean=504.47±0.66 sigma=54.58±0.49 500 500 mean=502.40±0.66 sigma=54.52±0.49 400 400 300 300 200 200 100 100 0 0 200 400 600 800 1000 420 440 460 480 500 520 540 560 580 600 amplitude amplitude • compared to the values without common mode both fjlters are almost completely eliminating the shift 10
with common mode efgect (4) signal peak amplitudes signal peak amplitudes count count mean=505.09±0.66 sigma=54.64±0.49 600 600 mean=504.80±0.66 sigma=54.72±0.50 500 500 mean=503.293±0.66 sigma=54.68±0.49 400 400 300 300 200 200 100 100 0 0 200 400 600 800 1000 420 440 460 480 500 520 540 560 580 600 amplitude amplitude • with baseline correction the mean value is restored similarly well by both fjlters 11
more realistic simulation charge rate function signal 6 10 × rate [Hz] amplitude [ADC] 700 3000 600 2500 500 400 2000 300 1500 200 100 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 1100 1200 1300 1400 1500 1600 1700 1800 1900 time [samples] time [samples] • charge rate is not constant during data taking • simulation with charge rate variation 1-3GHz • sine wave variation for simplicity and to stay within limits of BC2 thresholds 12
signifjcant difgerences signal peak amplitudes signal peak amplitudes count count mean=505.09±0.66 mean=505.09±0.66 sigma=54.64±0.49 sigma=54.64±0.49 600 600 mean=514.21±0.73 mean=504.80±0.66 sigma=54.72±0.50 sigma=57.40±0.56 500 500 mean=503.293±0.66 mean=502.90±0.67 sigma=54.25±0.50 sigma=54.68±0.49 400 400 300 300 200 200 100 100 0 0 0 200 400 600 800 1000 0 200 400 600 800 1000 amplitude amplitude • compared to result without charge rate variation, BC3 fjlter shows no signifjcant deviation • BC2 fjlter is hitting its limits 13
further steps • Adjustable slope range must be defjned for fjnalizing SAMPA fjlter design as bit width of registers must be specifjed prior to submiting the chip for production, this is time critical. • A better slope correction needs to be designed if shown deviations are a concern. This correction would be implemented in CRU/post processing, this is not time critical. • More realistic validation/simulation of the fjlters using AliRoot may be done next week when Marian/Mesut are available. 14
Questions? Comments? 15
Spares 16
Other simulation parameter ● Simulation statistics: 10000 simulated signal charges per histogram ● BC2 settings used: averaging length=8 samples upper=3 ADC lower=6 ADC ● BC3 settings used: upward=3/8 ADC/sample downward=6/8 ADC/sample 17
integer/fmoating point signal peak amplitudes signal peak amplitudes count count mean=504.90±0.67 mean=505.09±0.66 sigma=54.17±0.50 sigma=54.64±0.49 600 600 mean=514.21±0.73 mean=517.66±0.75 sigma=58.57±0.58 sigma=57.40±0.56 500 500 mean=502.44±0.67 mean=502.90±0.67 sigma=54.25±0.50 sigma=54.22±0.50 400 400 300 300 200 200 100 100 0 0 0 200 400 600 800 1000 0 200 400 600 800 1000 amplitude amplitude • simulation calculations done in fmoating point • integer simulation difgers only slightly 18
Recommend
More recommend
